release: v0.52.0 [skip ci]

Co-authored-by: emil User <emil.simeonov@tutanota.com>

.devin/skills/dependency-graph/SKILL.md

@@ -0,0 +1,135 @@
+# dependency-graph
+
+Map of the oblachno ecosystem. Knows which repo produces what, which
+repos depend on which, and the correct order for cross-repo changes.
+
+## When to Invoke
+
+Invoke this skill when:
+- Changes span multiple repos
+- A change in one repo requires version bumps in downstream repos
+- Deploying infrastructure that depends on published packages/images
+- Verifying the ecosystem is in a consistent state before deployment
+- Determining which repos to update and in what order
+
+## Prerequisites
+
+- All repos cloned under `/home/emo/dev/ideas/oblachno/`
+- `.env` with `DEVELOPER_GITEA_API_TOKEN` in each repo
+
+## Ecosystem Map
+
+```
+                    devx (PyPI package)
+                   /    |    \
+                  /     |     \
+            grm   sso-bridge   infra
+           (PyPI)  (PyPI+Docker) (deploys all)
+              |        |           |
+              v        v           v
+         infra bump  infra bump   staging
+         (auto PR)   (auto PR)   production
+                                      |
+                          mattermost-oidc (Docker image)
+                          (infra pulls :latest at deploy)
+```
+
+## Repositories
+
+| Repo | Produces | Consumers | Release Trigger |
+|------|----------|-----------|-----------------|
+| `devx` | PyPI package `devx` | grm, sso-bridge, infra | User-facing changes to `src/devx/**` |
+| `grm` | PyPI package `grm` | infra | User-facing changes to `src/grm/**` or `ansible/**` |
+| `sso-bridge` | PyPI package `sso_bridge` + Docker image | infra | User-facing changes to `src/sso_bridge/**` or `ansible/**` |
+| `infra` | Staging/production deployment | (end users) | User-facing changes + nightly gate |
+| `mattermost-oidc` | Docker image `mattermost-oidc` | infra (pulls at deploy) | `Dockerfile` or `build.yml` changes |
+
+## Dependency Chain
+
+### devx → all repos
+
+devx publishes to the Gitea PyPI registry. grm, sso-bridge, and infra
+pin devx in `pyproject.toml`:
+```toml
+"devx @ git+https://git.oblachno.oblachno.fyi/oblachno-oss/devx.git@vX.Y.Z"
+```
+
+When devx publishes a new version:
+1. grm, sso-bridge, and infra must bump their pinned devx version
+2. This is currently manual — no auto-dependency-PR from devx
+3. Each repo must `make setup` to pick up the new version
+
+### grm → infra
+
+grm publishes to PyPI. Its post-merge workflow auto-creates an infra
+dependency PR via `devx.ci.create_dependency_pr --repo oblachno/infra
+--package grm`. The PR bumps the pinned grm version in infra's
+`pyproject.toml`.
+
+### sso-bridge → infra
+
+sso-bridge publishes to PyPI AND builds a Docker image. Its post-merge
+workflow auto-creates an infra dependency PR via
+`devx.ci.create_dependency_pr --repo oblachno/infra --package
+sso_bridge`. The PR bumps the pinned sso_bridge version.
+
+The Docker image is pulled by infra at deploy time (`sso-bridge:latest`).
+
+### mattermost-oidc → infra
+
+mattermost-oidc builds a Docker image tagged `:latest` and `:MM_VERSION`.
+infra pulls `mattermost-oidc:latest` at deploy time. There is no
+auto-dependency-PR — infra simply pulls the latest image.
+
+### infra → staging/production
+
+infra deploys to staging and production. The deployment:
+1. Provisions VMs from golden images
+2. Runs Ansible roles (including grm and sso-bridge roles)
+3. Pulls Docker images (sso-bridge, mattermost-oidc)
+4. Configures services
+
+## Correct Order for Cross-Repo Changes
+
+When a change spans multiple repos, follow this order:
+
+1. **devx first** — if the change starts in devx, merge and publish devx
+   first. Wait for the PyPI publish job to complete.
+2. **Bump devx in consumers** — in grm/sso-bridge/infra, bump the pinned
+   devx version, run `make setup`, verify tests pass, merge.
+3. **grm/sso-bridge second** — merge and publish grm/sso-bridge. Wait
+   for the PyPI publish + Docker image build to complete.
+4. **Auto-dependency-PRs** — grm/sso-bridge post-merge auto-creates infra
+   PRs to bump pinned versions. Wait for these PRs to appear.
+5. **Merge infra dependency PRs** — review and merge the auto-created
+   infra PRs.
+6. **infra last** — deploy to staging, validate, promote to production.
+
+## State Verification Before Deployment
+
+Before deploying infra, verify:
+
+1. **devx version consistent** — all repos pin the same devx version
+2. **grm published** — latest grm tag exists in PyPI
+3. **sso-bridge published** — latest sso_bridge tag exists in PyPI
+4. **sso-bridge image built** — latest sso-bridge Docker image exists
+5. **mattermost-oidc image built** — latest mattermost-oidc image exists
+6. **infra pins match published versions** — no stale pins
+7. **Nightly gate green** — `NIGHTLY_STATUS` is not `failed`
+
+## Quick Check Commands
+
+```bash
+# Check latest devx version
+curl -sS https://git.oblachno.oblachno.fyi/api/v1/repos/oblachno-oss/devx/releases/latest | python3 -c "import json,sys; print(json.load(sys.stdin).get('tag_name','?'))"
+
+# Check pinned devx version in each repo
+for repo in grm sso-bridge infra; do
+  echo -n "$repo: "; grep 'devx @' /home/emo/dev/ideas/oblachno/$repo/pyproject.toml | grep -oP 'v[\d.]+'
+done
+
+# Check latest sso-bridge image build
+curl -sS -H "Authorization: token $DEVELOPER_GITEA_API_TOKEN" \
+  "https://git.oblachno.oblachno.fyi/api/v1/repos/oblachno/sso-bridge/actions/runs?per_page=5" \
+  | python3 -c "import json,sys; [print(r['id'],r['status'],r['conclusion']) for r in json.load(sys.stdin).get('workflow_runs',[]) if r.get('event')=='push']"
+```

.devin/skills/deployment-coordination/SKILL.md

@@ -0,0 +1,90 @@
+# deployment-coordination
+
+How devx releases propagate to downstream repos. devx is the base
+package — all other repos pin it. A devx release must complete before
+consumers can bump.
+
+## When to Invoke
+
+Invoke this skill when:
+- Changes to devx affect downstream repos (grm, sso-bridge, infra)
+- Preparing a devx release that other repos depend on
+- Verifying downstream repos have bumped to the latest devx version
+- Coordinating a multi-repo change that starts in devx
+
+## Prerequisites
+
+- devx repo at `/home/emo/dev/ideas/oblachno/devx`
+- `.env` with `DEVELOPER_GITEA_API_TOKEN`
+- See `dependency-graph` skill for the full ecosystem map
+
+## What devx Produces
+
+devx publishes a Python package to the Gitea PyPI registry. Downstream
+repos pin it:
+```toml
+"devx @ git+https://git.oblachno.oblachno.fyi/oblachno-oss/devx.git@vX.Y.Z"
+```
+
+## Release Flow
+
+1. PR merged to master
+2. Post-merge workflow runs `devx.ci.release` — classifies changes
+3. If user-facing changes: git-cliff bumps version, creates tag, pushes
+4. `devx.ci.publish` builds and publishes to Gitea PyPI registry
+5. Downstream repos must bump their pinned devx version
+
+## Downstream Consumers
+
+| Repo | Pin location | Auto-bump? |
+|------|-------------|------------|
+| grm | `pyproject.toml` | No — manual |
+| sso-bridge | `pyproject.toml` | No — manual |
+| infra | `pyproject.toml` | No — manual |
+
+devx does NOT auto-create dependency PRs in downstream repos. Bumping
+is manual: create a PR in each downstream repo to update the pinned
+version.
+
+## Coordinating a devx Change
+
+When a change to devx affects downstream repos:
+
+1. **Merge devx PR** — wait for post-merge publish to complete
+2. **Verify publish** — check the new tag exists:
+   ```bash
+   curl -sS -H "Authorization: token $DEVELOPER_GITEA_API_TOKEN" \
+     https://git.oblachno.oblachno.fyi/api/v1/repos/oblachno-oss/devx/releases/latest \
+     | python3 -c "import json,sys; print(json.load(sys.stdin).get('tag_name','?'))"
+   ```
+3. **Bump downstream repos** — for each repo (grm, sso-bridge, infra):
+   - Update `devx @ ...@vX.Y.Z` in `pyproject.toml`
+   - Run `make setup` to install the new version
+   - Run `make pytest-cov` to verify compatibility
+   - Create and merge a PR
+4. **Verify infra deploys** — after infra bumps, verify staging deploy
+   picks up the new devx version
+
+## State Verification
+
+Before starting a devx change, verify current state:
+```bash
+# Current devx version
+grep '__version__' /home/emo/dev/ideas/oblachno/devx/src/devx/__init__.py
+
+# What each repo pins
+for repo in grm sso-bridge infra; do
+  echo -n "$repo pins: "
+  grep 'devx @' /home/emo/dev/ideas/oblachno/$repo/pyproject.toml | grep -oP 'v[\d.]+'
+done
+```
+
+If pins are inconsistent across repos, bump them to the latest published
+version before starting new work.
+
+## Common Mistakes
+
+- Merging a devx PR and immediately merging downstream PRs without
+  waiting for the publish job to complete
+- Forgetting to bump infra (it has the most complex deploy pipeline)
+- Bumping only one downstream repo when the change affects all three

.devin/skills/devx-workflow/SKILL.md

@@ -2,11 +2,21 @@
 
 Quick reference for devx tools when working on the devx repo itself.
 
+## When to Invoke
+
+Invoke this skill when creating PRs, checking CI status, adding
+labels, rebasing branches, or performing any PR lifecycle operation.
+
+## Prerequisites
+
+- `.venv` exists (run `make setup` if not)
+- `.env` with `DEVELOPER_GITEA_API_TOKEN`, `VIKUNJA_TOKEN`
+
 ## PR Workflow (use these, not raw git/tea/MCP)
 
 | Task | Command |
 |------|---------|
-| Create Vikunja task | `make create-task -- --title "..." --description "..."` |
+| Create Vikunja task | `.venv/bin/python -m devx.tools.create_task --title "..." --description "..."` (make target doesn't forward args) |
 | Create PR | `make create-pr` |
 | Push + create PR | `make push-with-pr` |
 | Check CI status | `make devx-pr-status` or `make devx-pr-status PR=42 WAIT=1` |

.devin/skills/skill-creation/SKILL.md

@@ -0,0 +1,142 @@
+# skill-creation
+
+How to create, validate, and maintain Devin skills. Skills must be
+clear, succinct, and actionable — no AI slop.
+
+## When to Invoke
+
+Invoke this skill when:
+- Creating a new skill
+- Amending an existing skill
+- Evaluating whether a skill is needed
+- Reviewing a PR that adds or modifies skills
+
+## Prerequisites
+
+- Skill directory: `.devin/skills/<skill-name>/SKILL.md`
+- Validator: `tests/test_skills.py` (infra) or reference to it
+- Tests: `tests/unit/test_skills_validation.py` (infra)
+
+## When to Create a Skill
+
+Create a skill when:
+- An agent struggles with a task repeatedly (branch hygiene, PR order)
+- A workflow has non-obvious ordering constraints (deployment coordination)
+- A task requires specific tool usage over raw commands (CI monitoring)
+- Multiple agents need shared context (dependency graph)
+
+Do NOT create a skill for:
+- One-off tasks (use a spec instead)
+- Tasks already covered by AGENTS.md
+- Tasks that are obvious from the Makefile or README
+- Tasks that change frequently (skills should be stable)
+
+## Skill Structure
+
+Every skill MUST have:
+
+```markdown
+# <skill-name>
+
+One-line description of what the skill does.
+
+## When to Invoke
+
+2-4 bullet points describing when to use this skill.
+
+## Prerequisites
+
+What must exist before using the skill (venv, .env, tools).
+
+## <Core Content>
+
+The actual guidance. Keep it actionable.
+
+## Verification (if applicable)
+
+How to verify the skill's guidance works.
+
+## Common Mistakes (if applicable)
+
+What agents get wrong without this skill.
+```
+
+## Quality Standards
+
+### Do
+- **Be specific.** Reference exact make targets, file paths, commands.
+- **Be concise.** Each section should be scannable in under 30 seconds.
+- **Be actionable.** Every paragraph should tell the agent what to DO.
+- **Use tables** for command reference, mappings, and comparisons.
+- **Use code blocks** for commands the agent should run.
+- **Link to other skills** when related (e.g., "See `dependency-graph` skill").
+
+### Don't
+- **No preamble.** Don't start with "This skill helps agents..." — just state what it does.
+- **No filler.** Don't repeat information from AGENTS.md or other skills.
+- **No vague advice.** "Be careful with branches" is useless. "Run `git branch --show-current` before every commit" is useful.
+- **No AI slop.** Don't write "In this comprehensive guide, we will explore..." — just give the guidance.
+- **No redundant sections.** If "Common Mistakes" would repeat "When to Invoke", skip it.
+- **No marketing.** Don't describe the skill as "powerful" or "comprehensive".
+
+## Scope Rules
+
+- **One skill per concern.** Don't mix branch hygiene with CI monitoring.
+- **Project-specific, not generic.** Skills reference this repo's make targets, file paths, and conventions — not abstract advice.
+- **Shared skills must be identical across repos.** Use `SHARED_SKILLS` in the validator to enforce this.
+- **Per-repo skills must reflect that repo's reality.** Don't copy infra-specific targets to sso-bridge.
+
+## Automated Validation
+
+Every skill must pass the validator (`tests/test_skills.py`). The validator checks:
+
+1. **Structure** — H1 title, "When to Invoke" section, "Prerequisites" section
+2. **Commands** — referenced `make <target>` commands exist in Makefile or devx.mak
+3. **Paths** — referenced file paths exist in the repo
+4. **Shared skills** — identical content across repos (SHA-256 comparison)
+5. **No drift** — no references to nonexistent commands or files
+
+Run the validator:
+```bash
+python3 tests/test_skills.py --repo infra --repo sso-bridge
+```
+
+## Effectiveness Evaluation
+
+### Static Checks (automated, CI)
+
+The validator runs in CI as part of `make pytest-cov`. A failing skill
+test blocks the PR. This catches:
+- Missing sections
+- Invalid commands
+- Broken file references
+- Cross-repo drift
+
+### Runtime Metrics (manual, periodic)
+
+Track these signals to evaluate skill effectiveness:
+- **Skill invocation frequency** — how often agents invoke the skill
+- **Success rate when invoked** — did the skill prevent the mistake it targets?
+- **Feedback issues** — agents create Gitea issues with `feedback` label when a skill is unclear or wrong
+- **Mistake recurrence** — if agents still make the mistake the skill targets, the skill needs improvement
+
+### Retrospective Review
+
+Periodically (monthly or after major incidents) review skills:
+1. List all skills and their last-modified dates
+2. Check for feedback issues tagged `skill-improvement`
+3. Verify referenced commands still exist (run validator)
+4. Remove skills that are no longer relevant
+5. Update skills where mistakes still recur
+6. Document lessons in this skill's "Common Mistakes" section
+
+## Creating a New Skill — Checklist
+
+- [ ] Identify the repeated struggle or non-obvious workflow
+- [ ] Check no existing skill covers it
+- [ ] Write the skill following the structure above
+- [ ] Run `python3 tests/test_skills.py` — must pass
+- [ ] Run `make pytest-cov` — must pass with 100% coverage
+- [ ] If shared across repos, copy identical content to each repo
+- [ ] Add the skill to `SHARED_SKILLS` in the validator if shared
+- [ ] Create PR, verify CI passes, merge

.devin/skills/spec-driven-development/SKILL.md

@@ -1,5 +1,14 @@
 # Spec-Driven Development
 
+## When to Invoke
+
+Invoke this skill when starting any change — every PR requires a spec
+at `docs/specs/<TASK-ID>.md` that CI validates before merge.
+
+## Prerequisites
+
+- A Vikunja task ID (`DEVX-N`) — see `vikunja-tasks` skill
+
 ## Overview
 
 Every change starts with a spec. No spec, no code. No code, no PR.

.devin/skills/testing-and-debugging/SKILL.md

@@ -3,6 +3,17 @@
 Make targets for testing, debugging, and CI investigation. **Use these
 instead of raw `pytest`, `ruff`, or `actionlint` commands.**
 
+## When to Invoke
+
+Invoke this skill when running tests, investigating CI failures, or
+linting before push. Also invoke when asked to "run tests", "check
+coverage", or "debug a failure".
+
+## Prerequisites
+
+- `.venv` exists (run `make setup` if not)
+- Tools installed (run `make install-tools` for actionlint/act_runner)
+
 ## Why Make Targets
 
 Make targets encapsulate the correct venv activation, PYTHONPATH, env

.devin/skills/vikunja-tasks/SKILL.md

@@ -0,0 +1,74 @@
+# vikunja-tasks
+
+Vikunja task lifecycle beyond `create`: querying status, closing, and
+recovering when the tracker is unreachable.
+
+## When to Invoke
+
+- Creating, closing, or checking a Vikunja task
+- A spec workflow step needs the task ID or done state
+- `vikunja.oblachno.oblachno.fyi` fails to resolve / times out
+
+## Prerequisites
+
+- `.env` with `VIKUNJA_TOKEN`
+- Project ID comes from `[tool.devx]` in `pyproject.toml`
+  (`DEVX_VIKUNJA_PROJECT_ID`)
+
+## Create
+
+`make create-task` does **not** forward arguments — call the module:
+
+```bash
+.venv/bin/python -m devx.tools.create_task \
+  --title "Task title (no DEVX-N prefix)" \
+  --description "<h2>Context</h2><p>...</p>"
+```
+
+Prints `DEVX-N` + next steps. Title must not include the task-ID
+prefix (auto-merge prepends it; a manual prefix double-prefixes the
+PR title and fails validation).
+
+## Query / Close
+
+```bash
+# Task details (ID = numeric part of DEVX-N)
+curl -sf -H "Authorization: Bearer $VIKUNJA_TOKEN" \
+  "https://vikunja.oblachno.oblachno.fyi/api/v1/tasks/<N>"
+
+# Close: mark done
+curl -sf -X POST -H "Authorization: Bearer $VIKUNJA_TOKEN" \
+  -H "Content-Type: application/json" -d '{"done":true}' \
+  "https://vikunja.oblachno.oblachno.fyi/api/v1/tasks/<N>"
+```
+
+Post-merge automation marks the task done when the PR squash-merges —
+manual close is only needed for abandoned/superseded tasks.
+
+## Task-ID / Spec Collisions
+
+Vikunja IDs can collide with historical spec files (an old task reused
+the number). Convention: preserve the old file as
+`docs/specs/<ID>-<topic>-historical.md`, then write the new spec at
+`docs/specs/<ID>.md`. Check `git log` on the existing spec before
+moving it.
+
+## Tracker Unreachable
+
+If the Vikunja host fails DNS/TLS:
+
+1. Don't block the whole workflow — record the intended task title in
+   the spec draft and retry `create_task` before branching.
+2. Never invent an ID — branch/PR titles must match a real task or
+   `pre_push_check` / auto-merge validation fails.
+3. DNS failures observed so far were transient; retry after a few
+   minutes before escalating.
+
+## Common Mistakes
+
+- `make create-task -- --title ...` — args are dropped; use the module
+  call above (forwarding fix is S11 scope).
+- Including `DEVX-N:` in the task title — double prefix breaks
+  auto-merge.
+- Closing a task whose PR is still open — auto-merge's post-merge
+  step handles the close; manual close confuses the audit trail.

CHANGELOG.md

@@ -2,6 +2,54 @@
 
 All notable changes to this project will be documented in this file.
 
+## [0.52.0] - 2026-09-19
+
+### Features
+
+- *(ci)* Verify producer container artifact before opening dependency PR
+
+## [0.51.12] - 2026-09-19
+
+### Bug Fixes
+
+- Create_dependency_pr clones target repo instead of editing producer checkout
+
+## [0.51.11] - 2026-09-19
+
+### Bug Fixes
+
+- *(ci)* Retry Vikunja lookups and surface self-approval merge failures
+
+## [0.51.10] - 2026-09-17
+
+### Bug Fixes
+
+- Read Gitea repo-variable data field; fail closed on unknown nightly status
+
+## [0.51.9] - 2026-08-26
+
+### Bug Fixes
+
+- Exclude docs/plans/* from PR size check
+
+## [0.51.8] - 2026-08-26
+
+### Bug Fixes
+
+- Accept deps: as valid conventional commit type
+
+## [0.51.7] - 2026-08-26
+
+### Bug Fixes
+
+- Increase HTTP 500 retry count to 5 with longer backoff and visible logging
+
+## [0.51.6] - 2026-08-26
+
+### Bug Fixes
+
+- Use stderr=STDOUT to capture all docker push output in one stream
+
 ## [0.51.5] - 2026-08-26
 
 ### Bug Fixes

README.md

@@ -16,12 +16,12 @@ quality badges.
 
 [![CI](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions/workflows/ci.yml/badge.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
 [![License: GPL-3.0](https://img.shields.io/badge/license-GPL--3.0-blue)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/src/branch/master/LICENSE)
-[![Coverage](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/coverage.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Tests](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/tests.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Docs](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/docs.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/wiki)
-[![Code Quality](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/quality.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Version](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/version.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/releases)
-[![Python](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/python.svg)](https://www.python.org/downloads/)
+[![Coverage](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/coverage.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Tests](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/tests.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Docs](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/docs.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/wiki)
+[![Code Quality](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/quality.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Version](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/version.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/releases)
+[![Python](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/python.svg)](https://www.python.org/downloads/)
 
 ## Why devx?
 
@@ -87,7 +87,7 @@ extra index and list devx in your dependencies:
 ```toml
 [project]
 dependencies = [
-    "devx>=0.51.5",
+    "devx>=0.52.0",
 ]
 
 [tool.pip]
@@ -101,8 +101,8 @@ pip install -e .
 ```
 
 > **Note:** If your project requires a specific devx version, pin it in
-> `dependencies` (for example, `"devx==0.51.5"`) or use a version constraint
-> (for example, `"devx>=0.51.5,<0.52"`).
+> `dependencies` (for example, `"devx==0.52.0"`) or use a version constraint
+> (for example, `"devx>=0.52.0,<0.53"`).
 
 ### Optional extras
 

docs/index.md

@@ -12,12 +12,12 @@ project to be reusable across all oblachno-oss repositories.
 
 [![CI](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions/workflows/ci.yml/badge.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
 [![License: GPL-3.0](https://img.shields.io/badge/license-GPL--3.0-blue)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/src/branch/master/LICENSE)
-[![Coverage](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/coverage.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Tests](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/tests.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Docs](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/docs.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/wiki)
-[![Code Quality](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/quality.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
-[![Version](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/version.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/releases)
-[![Python](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/7660d5019b904710f01bf56fc45e0b9d4d87b53d/python.svg)](https://www.python.org/downloads/)
+[![Coverage](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/coverage.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Tests](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/tests.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Docs](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/docs.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/wiki)
+[![Code Quality](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/quality.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/actions)
+[![Version](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/version.svg)](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/releases)
+[![Python](https://git.oblachno.oblachno.fyi/oblachno-oss/devx/raw/commit/9f6cf5738cdf4edb7be7691a0bd7a4c4584ff9a2/python.svg)](https://www.python.org/downloads/)
 
 ## Overview
 
@@ -74,14 +74,14 @@ Add devx to your `pyproject.toml` dependencies and configure the registry:
 ```toml
 [project]
 dependencies = [
-    "devx>=0.51.5",
+    "devx>=0.52.0",
 ]
 
 [tool.pip]
 extra-index-url = "https://git.oblachno.oblachno.fyi/api/packages/oblachno-oss/pypi/simple"
 ```
 
-Pin a specific version if needed: `"devx==0.51.5"` or `"devx>=0.51.5,<0.52"`.
+Pin a specific version if needed: `"devx==0.52.0"` or `"devx>=0.52.0,<0.53"`.
 
 ### Optional extras
 

docs/specs/DEVX-162-registry-push-race-historical.md

@@ -0,0 +1,41 @@
+# DEVX-162: Fix registry push race condition: serialize uploads + retry on HTTP 500
+
+## Problem
+The Gitea container registry (v1.27.2) has a known race condition in
+`BlobUploader.Append()` where concurrent blob uploads cause the file
+offset and DB model to get out of sync, producing HTTP 500 "offset
+mismatch between file and model" errors. This causes the build-images
+workflow to fail intermittently when pushing runner images.
+
+The `package_blob_upload` table accumulates stale entries from failed
+uploads that worsen the problem over time.
+
+## Approach
+Two fixes in devx (a third fix — scheduled cleanup — is tracked
+separately as OBL-INFRA-537):
+
+1. Set `DOCKER_MAX_CONCURRENT_UPLOADS=1` in the build-images workflow
+   to serialize blob uploads and avoid the race condition.
+
+2. Add HTTP 500 retry logic to `push_image` in `build_image.py`.
+   When a push fails with HTTP 500 (not "already exists"), retry up
+   to 3 times with exponential backoff (5s, 10s, 20s).
+
+REQ-1: Build-images workflow sets DOCKER_MAX_CONCURRENT_UPLOADS=1
+REQ-2: push_image retries on HTTP 500 with exponential backoff
+REQ-3: All existing tests pass with 100% coverage
+
+## Test Plan
+- Unit tests for retry logic (mock subprocess)
+- Manual: trigger build-images workflow and verify push succeeds
+
+## Deploy Plan
+- Merge to master
+
+## Rollback Plan
+- Revert the merge commit
+
+## Acceptance Criteria
+- [x] REQ-1: Build-images workflow sets DOCKER_MAX_CONCURRENT_UPLOADS=1
+- [x] REQ-2: push_image retries on HTTP 500 with exponential backoff
+- [x] REQ-3: All existing tests pass with 100% coverage

docs/specs/DEVX-162.md

@@ -1,41 +1,58 @@
-# DEVX-162: Fix registry push race condition: serialize uploads + retry on HTTP 500
+# DEVX-162: Fix Gitea repo-variable read contract and fail closed on unknown nightly status
 
 ## Problem
-The Gitea container registry (v1.27.2) has a known race condition in
-`BlobUploader.Append()` where concurrent blob uploads cause the file
-offset and DB model to get out of sync, producing HTTP 500 "offset
-mismatch between file and model" errors. This causes the build-images
-workflow to fail intermittently when pushing runner images.
 
-The `package_blob_upload` table accumulates stale entries from failed
-uploads that worsen the problem over time.
+`GiteaClient.get_repo_variable()` reads `body["value"]`, but the deployed
+Gitea returns the variable payload in the `data` field:
+
+```json
+{"owner_id":0,"repo_id":1,"name":"NIGHTLY_STATUS","data":"passed:5842","description":""}
+```
+
+Every read therefore returns `None`. `devx.ci.nightly_gate --action check`
+interprets `None` as "bootstrap — allow deploy," so a real `failed:<run>` status
+is invisible and the gate is permanently fail-open. Infra nightly run 5800 set
+`NIGHTLY_STATUS=passed:5800` while platform/customer integration tests were
+still failing — and even a correct `failed` value would have been ignored.
+
+Additionally, an unrecognized non-empty status currently allows deploys
+(fail-open instead of fail-closed).
+
+Verified against the live API: `POST`/`PUT` accept `{"value": ...}` and work;
+only the GET response uses `data`. The earlier `DEVX-162` spec (registry push
+race) is preserved as `DEVX-162-registry-push-race-historical.md`.
 
 ## Approach
-Two fixes in devx (a third fix — scheduled cleanup — is tracked
-separately as OBL-INFRA-537):
 
-1. Set `DOCKER_MAX_CONCURRENT_UPLOADS=1` in the build-images workflow
-   to serialize blob uploads and avoid the race condition.
+REQ-1: `src/devx/api_clients.py` — `get_repo_variable` reads `data` first
+and falls back to `value` for older server/fixture compatibility. Write
+path unchanged (PUT/POST `{"value": ...}` verified live: 201/204).
 
-2. Add HTTP 500 retry logic to `push_image` in `build_image.py`.
-   When a push fails with HTTP 500 (not "already exists"), retry up
-   to 3 times with exponential backoff (5s, 10s, 20s).
+REQ-2: `src/devx/ci/nightly_gate.py` — unknown non-empty status blocks the
+deploy (fail closed) instead of allowing it. Unset (bootstrap) still
+allows.
 
-REQ-1: Build-images workflow sets DOCKER_MAX_CONCURRENT_UPLOADS=1
-REQ-2: push_image retries on HTTP 500 with exponential backoff
-REQ-3: All existing tests pass with 100% coverage
+REQ-3: Tests cover the `data` field, the `value` fallback, and fail-closed
+unknown status.
 
 ## Test Plan
-- Unit tests for retry logic (mock subprocess)
-- Manual: trigger build-images workflow and verify push succeeds
+
+- `pytest tests/unit/test_api_clients.py tests/unit/test_nightly_gate.py`
+- `make lint-all` (ruff, pyright, bandit, translations)
 
 ## Deploy Plan
-- Merge to master
+
+Merge via auto-merge; post-merge workflow publishes a new devx package to the
+Gitea PyPI registry and opens the infra dependency-bump PR automatically.
 
 ## Rollback Plan
-- Revert the merge commit
+
+Revert the commit; infra's pinned devx version keeps the previous behavior
+until the dependency PR lands.
 
 ## Acceptance Criteria
-- [x] REQ-1: Build-images workflow sets DOCKER_MAX_CONCURRENT_UPLOADS=1
-- [x] REQ-2: push_image retries on HTTP 500 with exponential backoff
-- [x] REQ-3: All existing tests pass with 100% coverage
+
+- [x] `get_repo_variable` returns the `data` field and falls back to `value`.
+- [x] Unknown nightly status exits non-zero and writes `nightly-gate-passed=false`.
+- [x] Unit tests cover `data`, `value` fallback and fail-closed unknown status.
+- [x] `make lint-all` and unit tests pass.

docs/specs/DEVX-163-run-push-stdout-historical.md

@@ -0,0 +1,33 @@
+# DEVX-163: Fix _run_push to check stdout for HTTP 500
+
+## Problem
+`_run_push` only checked `result.stderr` for HTTP 500, but docker push
+sends the "received unexpected HTTP status: 500 Internal Server Error"
+message to **stdout**, not stderr. This means the tenacity retry logic
+added in DEVX-162 never triggered — the push failed immediately without
+retrying.
+
+## Approach
+Check both `result.stdout` and `result.stderr` for the "500" status code.
+Also update the "already exists" check in `push_image` to check both
+streams, since docker may send that message to stdout as well.
+
+REQ-1: _run_push checks both stdout and stderr for HTTP 500
+REQ-2: push_image "already exists" check uses combined stdout+stderr
+REQ-3: All existing tests pass with 100% coverage
+
+## Test Plan
+- Unit tests for stdout 500 detection
+- Unit tests for stderr 500 detection
+- Manual: trigger build-images workflow and verify retry works
+
+## Deploy Plan
+- Merge to master
+
+## Rollback Plan
+- Revert the merge commit
+
+## Acceptance Criteria
+- [x] REQ-1: _run_push checks both stdout and stderr for HTTP 500
+- [x] REQ-2: push_image "already exists" check uses combined stdout+stderr
+- [x] REQ-3: All existing tests pass with 100% coverage

docs/specs/DEVX-163.md

@@ -1,33 +1,46 @@
-# DEVX-163: Fix _run_push to check stdout for HTTP 500
+# DEVX-163: Add vikunja-tasks skill and skill validation tests, fix create-task docs
 
 ## Problem
-`_run_push` only checked `result.stderr` for HTTP 500, but docker push
-sends the "received unexpected HTTP status: 500 Internal Server Error"
-message to **stdout**, not stderr. This means the tenacity retry logic
-added in DEVX-162 never triggered — the push failed immediately without
-retrying.
+
+The OBL-INFRA-548 programme audit found devx lacks a Vikunja
+task-lifecycle skill and has no skill validation tests (infra and
+sso-bridge have them; grm gained them under GRM-171).
+`devx-workflow` documents `make create-task -- --title`, which fails
+because `devx-create-task` forwards no arguments.
 
 ## Approach
-Check both `result.stdout` and `result.stderr` for the "500" status code.
-Also update the "already exists" check in `push_image` to check both
-streams, since docker may send that message to stdout as well.
 
-REQ-1: _run_push checks both stdout and stderr for HTTP 500
-REQ-2: push_image "already exists" check uses combined stdout+stderr
-REQ-3: All existing tests pass with 100% coverage
+REQ-1: Add `vikunja-tasks` skill: create via module call, query,
+  close, spec-collision convention, unreachable-tracker handling.
+REQ-2: Add `tests/unit/test_skills_validation.py` covering
+  structure, make-target, file-ref checks + existence tests for all
+  skills.
+REQ-3: Fix broken `make create-task -- --title` documentation in
+  `devx-workflow` skill; add missing When to Invoke / Prerequisites
+  sections to older-format skills.
+
+Preserve the colliding spec as
+[DEVX-163-run-push-stdout-historical](DEVX-163-run-push-stdout-historical.md).
 
 ## Test Plan
-- Unit tests for stdout 500 detection
-- Unit tests for stderr 500 detection
-- Manual: trigger build-images workflow and verify retry works
+
+- `pytest tests/unit/test_skills_validation.py` passes (10 tests).
 
 ## Deploy Plan
-- Merge to master
+
+Documentation/skills only — auto-merge to master; no runtime deploy.
 
 ## Rollback Plan
-- Revert the merge commit
+
+Revert the squash-merge commit; skills are inert documentation.
 
 ## Acceptance Criteria
-- [x] REQ-1: _run_push checks both stdout and stderr for HTTP 500
-- [x] REQ-2: push_image "already exists" check uses combined stdout+stderr
-- [x] REQ-3: All existing tests pass with 100% coverage
+
+- [x] REQ-1: `vikunja-tasks` skill exists.
+- [x] REQ-2: `tests/unit/test_skills_validation.py` exists and passes.
+- [x] REQ-3: create-task docs corrected.
+
+## Out of Scope
+
+- Fixing `devx-create-task` argument forwarding (S11 backlog: the
+  devx.mak target takes no args; needs env-var or arg forwarding).

docs/specs/DEVX-164-docker-push-retry-historical.md

@@ -0,0 +1,34 @@
+# DEVX-164: Increase HTTP 500 retry count and backoff for docker push
+
+## Problem
+The HTTP 500 retry logic (DEVX-162, DEVX-163) works correctly — 3 retry
+attempts are made. But all 3 attempts fail because the Gitea registry's
+"offset mismatch" race condition needs more than ~15s to recover. The
+current backoff is 5s-20s with 3 attempts (total ~15s of waiting).
+
+## Approach
+Increase retry count from 3 to 5 and backoff from 5-20s to 10-60s,
+giving the registry up to ~2 minutes to recover. Add visible logging
+between retry attempts so the CI logs show the retry happening.
+
+REQ-1: Increase retry count from 3 to 5
+REQ-2: Increase backoff from 5-20s to 10-60s exponential
+REQ-3: Add visible logging between retry attempts (click.echo)
+REQ-4: All tests pass with 100% coverage
+
+## Test Plan
+- Unit tests verify retry count and backoff parameters
+- Unit tests verify logging output on retry
+- Manual: trigger build-images workflow and verify retries visible in logs
+
+## Deploy Plan
+- Merge to master
+
+## Rollback Plan
+- Revert the merge commit
+
+## Acceptance Criteria
+- [x] REQ-1: Increase retry count from 3 to 5
+- [x] REQ-2: Increase backoff from 5-20s to 10-60s exponential
+- [x] REQ-3: Add visible logging between retry attempts (click.echo)
+- [x] REQ-4: All tests pass with 100% coverage

docs/specs/DEVX-164.md

@@ -0,0 +1,57 @@
+# DEVX-164: auto-merge resilience — self-approval and Vikunja outage handling
+
+## Problem
+
+Two defects hit auto-merge during S02 work:
+
+1. `get_vikunja_task_title` crashes on transient Vikunja errors. The
+   Vikunja API returned 404/502 during a restart window (run 6093);
+   `list_project_tasks` treats 4xx as non-retryable `APIError`, so the
+   job failed immediately instead of riding out a short outage.
+2. When a PR author and the workflow's reviewer token map to the same
+   Gitea user, the auto-approve step is rejected ("approve your own
+   pull is not allowed") and the merge fails `405: not enough
+   approvals`. The generic merge error gives no hint that an external
+   approval is the fix (hit on sso-bridge #18, #19, and infra #1647's
+   approvals-only failure mode).
+
+## Approach
+
+REQ-1: Wrap the task-list pagination in `get_vikunja_task_title` with a
+bounded retry (tenacity, ~4 attempts, exponential backoff) covering
+`APIError` and `requests.RequestException`. A genuinely missing task
+still ends in the same "Could not find" ClickException.
+
+REQ-2: On merge `HTTP 405`, fetch PR reviews; when zero `APPROVED`
+reviews exist, extend the error with the self-approval explanation and
+the remediation (approve via a non-author account).
+
+REQ-3: Regression tests for both behaviors.
+
+## Files Affected
+
+- `src/devx/ci/auto_merge.py`
+- `tests/unit/test_auto_merge.py`
+
+## Test Plan
+
+- New tests: retry-then-success on transient APIError; retry-exhaustion
+  still raises; missing task still raises; 405 error includes
+  approvals diagnostic.
+- `make pytest-cov`, `make lint-all`.
+
+## Deploy Plan
+
+- Merge → next release publishes the package; consuming repos pick it
+  up on their next CI run (devx is pinned per-repo, bump via the usual
+  dependency PR flow).
+
+## Rollback Plan
+
+- Revert; previous behavior returns.
+
+## Acceptance Criteria
+
+- [x] REQ-1: Vikunja task-list retries transient API failures
+- [x] REQ-2: 405 merge error reports approval state + self-approval hint
+- [x] REQ-3: Regression tests added and passing

docs/specs/DEVX-165.md

@@ -0,0 +1,31 @@
+# DEVX-165: Accept deps: as valid conventional commit type
+
+## Problem
+The commit validator rejects `deps:` as a conventional commit type, causing
+post-merge CI failures on grm and sso-bridge repos where automated dependency
+bump PRs use `deps: bump devx...` as the commit message.
+
+## Approach
+REQ-1: Add `deps` to `CONVENTIONAL_RE` in `src/devx/config.py`
+REQ-2: Update the allowed types list in the error message in
+  `src/devx/ci/validate_commit_msg.py`
+REQ-3: Add test coverage for `deps:` type in `tests/unit/test_config.py`
+  and `tests/unit/test_validate_commit_msg.py`
+
+## Test Plan
+- `make pytest-cov` passes with 100% coverage
+- `make lint-all` passes
+
+## Deploy Plan
+- Merge to master → post-merge auto-publishes new devx version
+- grm and sso-bridge bump devx version to pick up the fix
+
+## Rollback Plan
+- Revert the merge commit
+
+## Acceptance Criteria
+- [x] REQ-1: Add `deps` to `CONVENTIONAL_RE` in `src/devx/config.py`
+- [x] REQ-2: Update the allowed types list in the error message in
+  `src/devx/ci/validate_commit_msg.py`
+- [x] REQ-3: Add test coverage for `deps:` type in `tests/unit/test_config.py`
+  and `tests/unit/test_validate_commit_msg.py`

docs/specs/DEVX-166.md

@@ -0,0 +1,47 @@
+# DEVX-166: create_dependency_pr must clone the target repo
+
+## Problem
+
+`create_dependency_pr` resolves the pinned-version file and runs all
+git operations in the current working directory. Producer post-merge
+workflows (grm, sso-bridge) invoke it from the *producer* checkout, so
+it searches/modifies the wrong repository: `find_pinned_version` reads
+files that do not exist there, and the git fetch/checkout/commit/push
+sequence runs in the producer clone. The failure is silent — producer
+workflows append `|| echo warning`, so a no-op looks like success.
+
+## Approach
+
+REQ-1: Clone the target repo (`--repo`) into a temporary directory with
+an authenticated `http.extraHeader`, then run every file lookup and git
+operation (fetch, checkout, add, commit, push) inside that clone. The
+push uses the same auth header config.
+
+REQ-2: Tests mock `subprocess.run` so no real clone happens in the unit
+suite (test-isolation gate).
+
+## Files Affected
+
+- `src/devx/ci/create_dependency_pr.py`
+- `tests/unit/test_create_dependency_pr.py`
+
+## Test Plan
+
+- Existing CLI tests keep passing with the subprocess mock in place.
+- Verify the clone command targets the `--repo` URL and that git ops
+  run with `cwd=<clone>` (asserted via the mock's call list).
+
+## Deploy Plan
+
+Merge via auto-merge after green CI. The fix takes effect the next time
+a producer post-merge workflow invokes `create_dependency_pr`.
+
+## Rollback Plan
+
+Revert the squash-merge commit on master; the previous (broken) CWD
+behavior returns, which is strictly worse — no state is created.
+
+## Acceptance Criteria
+
+- [x] REQ-1: target repo cloned to tempdir; all file/git ops run in the clone
+- [x] REQ-2: unit tests never spawn a real git subprocess

docs/specs/DEVX-167-skills-historical.md

@@ -0,0 +1,64 @@
+# DEVX-167: Add dependency-graph, deployment-coordination, and skill-creation skills
+
+## Problem
+Agents working across the oblachno ecosystem lack shared, persistent
+context for three recurring pain points:
+
+1. **Cross-repo dependency ordering** — agents frequently merge
+   downstream PRs before the upstream publish job completes, or forget
+   to bump infra. There is no single reference for which repo produces
+   what and in what order changes must propagate.
+2. **devx release coordination** — devx is the base package pinned by
+   grm, sso-bridge, and infra. Agents repeatedly merge devx PRs and
+   immediately merge downstream bumps without waiting for the PyPI
+   publish job, or bump only one consumer when a change affects all
+   three.
+3. **Skill quality drift** — skills are created ad hoc with inconsistent
+   structure, vague advice, and no automated validation reference. New
+   skills miss required sections, reference nonexistent make targets,
+   and drift across repos.
+
+## Approach
+Add three SKILL.md files under `.devin/skills/`:
+
+REQ-1: `dependency-graph` — shared skill mapping the oblachno ecosystem
+(repos, what each produces, consumers, release triggers, correct
+cross-repo change order, state verification checklist)
+
+REQ-2: `deployment-coordination` — devx-specific skill covering the
+devx release flow, downstream consumers, manual bump procedure, and
+common mistakes when coordinating a devx change
+
+REQ-3: `skill-creation` — shared skill defining skill structure,
+quality standards, scope rules, automated validation reference, and a
+creation checklist
+
+## Files Affected
+- `.devin/skills/dependency-graph/SKILL.md` (new)
+- `.devin/skills/deployment-coordination/SKILL.md` (new)
+- `.devin/skills/skill-creation/SKILL.md` (new)
+- `docs/specs/DEVX-167.md` (new)
+
+## Test Plan
+- Verify all three SKILL.md files follow the required structure (H1
+  title, When to Invoke, Prerequisites sections)
+- Verify referenced make targets and file paths exist
+- Run `make pytest-cov` — skill validation tests must pass
+
+## Deploy Plan
+- Merge to master; skills are consumed by agents immediately on next
+  invocation — no build or deploy step required
+
+## Rollback Plan
+- Revert the merge commit; remove the three skill directories
+
+## Acceptance Criteria
+- [x] REQ-1: dependency-graph skill exists with ecosystem map, repo
+      table, dependency chain, cross-repo change order, and state
+      verification checklist
+- [x] REQ-2: deployment-coordination skill exists with devx release
+      flow, downstream consumer table, coordination steps, and common
+      mistakes
+- [x] REQ-3: skill-creation skill exists with structure template,
+      quality standards, scope rules, validation reference, and
+      creation checklist

docs/specs/DEVX-167.md

@@ -0,0 +1,57 @@
+# DEVX-167: S03 artifact integrity — verify producer artifact + cleanup protection
+
+## Problem
+
+S03 (OBL-INFRA-548 REQ-3) requires that dependency PRs only open after
+the producer artifact exists and is content-addressable, and that
+registry cleanup never deletes a version pinned by a release manifest.
+Two gaps:
+
+1. `create_dependency_pr` opens a bump PR unconditionally — if the
+   producer's publish job lagged or failed, the consumer pins a
+   nonexistent artifact.
+2. The sso-bridge image tag is derived from `__init__.py.__version__`,
+   which does not always equal the release git tag, so the tag to
+   verify must be decoupled from `--new-version`.
+
+## Approach
+
+REQ-1: `create_dependency_pr` gains `--verify-container <owner/name>`
+and `--container-tag <tag>`: before any branch/PR work it resolves the
+OCI digest of the image tag via the Gitea packages API (`manifest.json`
+blob sha256) and refuses the PR when the artifact is missing or
+unreadable. `--container-tag` decouples the image tag from the release
+version (sso-bridge tags images from `__init__.py.__version__`, not the
+git tag).
+
+REQ-2: Regression tests cover digest resolution, verification-failure
+aborts, invalid container format, and the `--container-tag` override.
+
+## Files Affected
+
+- `src/devx/ci/create_dependency_pr.py`
+- `src/devx/translations.json`
+- `tests/unit/test_create_dependency_pr.py`
+
+## Test Plan
+
+- Unit tests for `resolve_container_digest` (digest from manifest blob,
+  missing tag, missing blob, connection error).
+- CLI tests: verify runs before version lookup, digest resolution,
+  invalid format rejection, container-tag override.
+
+## Deploy Plan
+
+Merge via auto-merge after green CI. Producer post-merge workflows adopt
+the new flags in their own PRs (sso-bridge SSO-22 already passes them).
+
+## Rollback Plan
+
+Revert the squash-merge commit; the new options disappear and callers
+without them behave exactly as before.
+
+## Acceptance Criteria
+
+- [x] REQ-1: pre-PR OCI digest verification with --verify-container/--container-tag
+- [x] REQ-2: regression tests for all new behavior
+

docs/user/getting-started.md

@@ -48,12 +48,12 @@ Add devx to your `pyproject.toml`:
 ```toml
 [project]
 dependencies = [
-    "devx>=0.51.5",
+    "devx>=0.52.0",
 ]
 
 [project.optional-dependencies]
 dev = [
-    "devx>=0.51.5",
+    "devx>=0.52.0",
 ]
 ```
 

src/devx/__init__.py

@@ -6,4 +6,4 @@ create_dependency_pr, auto_merge, release, publish), developer tooling
 molecule testing helpers for Ansible projects.
 """
 
-__version__ = "0.51.5"
+__version__ = "0.52.0"

src/devx/api_clients.py

@@ -392,7 +392,10 @@ class GiteaClient:
         """
         try:
             r = self._request("GET", f"/actions/variables/{name}")
-            return r.json().get("value")
+            body = r.json()
+            if "data" in body:
+                return body["data"]
+            return body.get("value")
         except APIError as e:
             if e.status == 404:
                 return None

src/devx/ci/auto_merge.py

@@ -21,10 +21,12 @@ Usage:
 """
 
 import re
+import time
 from pathlib import Path
 from typing import Any
 
 import click
+import requests
 from dotenv import load_dotenv  # pyright: ignore[reportMissingImports,reportUnknownVariableType]
 
 from devx.api_clients import GiteaClient, VikunjaClient
@@ -110,6 +112,27 @@ def validate_pr_title(pr_title: str, task_id: str) -> None:
         )
 
 
+_VIKUNJA_LOOKUP_ATTEMPTS = 4
+_VIKUNJA_LOOKUP_BACKOFF = 5.0
+
+
+def _list_project_tasks(client: VikunjaClient, page: int) -> list[dict[str, Any]]:
+    """List Vikunja tasks with bounded retries for transient outages.
+
+    Implements REQ-1: during a Vikunja restart the tasks endpoint can briefly
+    return 404/502; retry a few times so title validation rides out the window
+    instead of stranding an otherwise-valid PR.
+    """
+    for attempt in range(1, _VIKUNJA_LOOKUP_ATTEMPTS + 1):
+        try:
+            return list(client.list_project_tasks(VIKUNJA_PROJECT_ID, page=page, per_page=DEFAULT_PER_PAGE))
+        except (APIError, requests.RequestException):
+            if attempt == _VIKUNJA_LOOKUP_ATTEMPTS:
+                raise
+            time.sleep(_VIKUNJA_LOOKUP_BACKOFF * attempt)
+    raise AssertionError("unreachable")  # pragma: no cover
+
+
 def get_vikunja_task_title(task_id: str) -> str:
     """Fetch the Vikunja task title for the given DEVX-N identifier.
 
@@ -124,7 +147,7 @@ def get_vikunja_task_title(task_id: str) -> str:
     client = VikunjaClient(VIKUNJA_API_URL, token)
     page = 1
     while True:
-        tasks = client.list_project_tasks(VIKUNJA_PROJECT_ID, page=page, per_page=DEFAULT_PER_PAGE)
+        tasks = _list_project_tasks(client, page)
         if not tasks:
             break
         matches = [t for t in tasks if t.get("identifier") == task_id]
@@ -272,14 +295,28 @@ def main(branch: str, pr_title: str, repo: str, pr_number: str) -> None:
             # Exit cleanly — the rebase triggers a new CI run that will retry.
             return
         else:
-            raise click.ClickException(
-                _(
-                    "Merge failed with HTTP {status}: {message}\n"
-                    "Please check the PR is ready and you have merge rights.",
-                    status=e.status,
-                    message=e.message,
-                )
-            ) from None
+            hint = ""
+            if e.status == 405:
+                # Implements: REQ-2 — surface approval state. When the PR author
+                # and the CI reviewer token map to the same Gitea user,
+                # self-approval is rejected and the merge fails 405.
+                try:
+                    reviews = client.get_pr_reviews(pr_num)
+                    if not any(r.get("state") == "APPROVED" for r in reviews):
+                        hint = _(
+                            "\nNo APPROVED review found on the PR. If the PR author and the"
+                            " CI reviewer token map to the same Gitea user, self-approval is"
+                            " rejected — approve the PR via a non-author account, then"
+                            " re-run the auto-merge job."
+                        )
+                except APIError:
+                    pass
+            msg = _(
+                "Merge failed with HTTP {status}: {message}\nPlease check the PR is ready and you have merge rights.",
+                status=e.status,
+                message=e.message,
+            )
+            raise click.ClickException(msg + hint) from None
 
     click.echo(
         _(

src/devx/ci/check_pr_size.py

@@ -36,6 +36,7 @@ DEFAULT_EXCLUDED_PATTERNS = [
     "CHANGELOG.md",
     "README.md",
     "docs/index.md",
+    "docs/plans/*",
     "*.svg",
     "uv.lock",
     "poetry.lock",

src/devx/ci/create_dependency_pr.py

@@ -21,9 +21,11 @@ from __future__ import annotations
 
 import re
 import subprocess  # nosec B404
+import tempfile
 from pathlib import Path
 
 import click
+import requests
 from dotenv import load_dotenv
 
 from devx.api_clients import GiteaClient
@@ -92,6 +94,55 @@ def update_pinned_version(file_path: str, package: str, old_version: str, new_ve
     return changed
 
 
+def resolve_container_digest(api_url: str, owner: str, name: str, tag: str, token: str) -> str:
+    """Resolve the OCI digest for a container image tag via the packages API.
+
+    Implements REQ-1: dependency PRs must only be opened after the producer
+    artifact exists — this raises ClickException when the tag is missing or
+    the registry call fails, so the PR is never opened against an artifact
+    that has not been published. The sha256 of the stored ``manifest.json``
+    blob is the manifest content digest (what ``docker pull`` reports).
+    """
+    url = f"{api_url}/packages/{owner}/container/{name}/{tag}/files"
+    headers = {"Authorization": f"token {token}"}
+    try:
+        resp = requests.get(url, headers=headers, timeout=30)  # nosec B310
+        resp.raise_for_status()
+    except requests.HTTPError as e:
+        status = e.response.status_code if e.response is not None else "?"
+        raise click.ClickException(
+            _(
+                "Container artifact {owner}/{name}:{tag} not found or unreadable (HTTP {status}). "
+                "Refusing to open a dependency PR for an unpublished artifact.",
+                owner=owner,
+                name=name,
+                tag=tag,
+                status=status,
+            )
+        ) from e
+    except requests.RequestException as e:
+        raise click.ClickException(
+            _(
+                "Registry lookup failed for {owner}/{name}:{tag}: {error}",
+                owner=owner,
+                name=name,
+                tag=tag,
+                error=e,
+            )
+        ) from e
+    for f in resp.json():
+        if f.get("name") == "manifest.json" and f.get("sha256"):
+            return f"sha256:{f['sha256']}"
+    raise click.ClickException(
+        _(
+            "Registry returned no manifest blob for {owner}/{name}:{tag}.",
+            owner=owner,
+            name=name,
+            tag=tag,
+        )
+    )
+
+
 def create_vikunja_task(title: str, description: str) -> str | None:
     """Create a Vikunja task and return its identifier (e.g., OBL-INFRA-531)."""
     try:
@@ -112,6 +163,24 @@ def create_vikunja_task(title: str, description: str) -> str | None:
 @click.option("--new-version", required=True, help=_("New version to pin"))
 @click.option("--source-repo", required=True, help=_("Source repo that published (owner/name)"))
 @click.option("--source-run-id", default="", help=_("CI run ID that triggered the publish"))
+@click.option(
+    "--verify-container",
+    default="",
+    help=_(
+        "Container to verify before opening the PR (owner/name). Resolves the OCI "
+        "digest of the tag matching --new-version (or --container-tag); the PR is "
+        "refused when the artifact is missing or unreadable."
+    ),
+)
+@click.option(
+    "--container-tag",
+    default="",
+    help=_(
+        "Container tag to verify with --verify-container (default: --new-version). "
+        "Use when the image tag differs from the release version, e.g. a package "
+        "__version__ tag vs a release git tag."
+    ),
+)
 @click.option("--dry-run", is_flag=True, default=False, help=_("Show what would be done without creating PR"))
 def cli(
     repo: str,
@@ -119,6 +188,8 @@ def cli(
     new_version: str,
     source_repo: str,
     source_run_id: str,
+    verify_container: str,
+    container_tag: str,
     dry_run: bool,
 ) -> None:
     """Create an infra PR to bump a pinned dependency version."""
@@ -128,11 +199,43 @@ def cli(
     owner, repo_name = repo.split("/", 1)
     client = GiteaClient(GITEA_API_URL, token, owner, repo_name)
 
+    # Implements: REQ-1 — verify the producer artifact exists and resolve its
+    # digest before any branch/PR work begins.
+    image_digest = ""
+    if verify_container:
+        if "/" not in verify_container:
+            raise click.ClickException(
+                _("Invalid container format: {container} (expected owner/name)", container=verify_container)
+            )
+        c_owner, c_name = verify_container.split("/", 1)
+        image_tag = container_tag or new_version
+        image_digest = resolve_container_digest(GITEA_API_URL, c_owner, c_name, image_tag, token)
+        click.echo(
+            _(
+                "[dep-pr] Verified {container}:{version} -> {digest}",
+                container=verify_container,
+                version=image_tag,
+                digest=image_digest,
+            )
+        )
+
+    # Clone the target repo — this tool runs from the *producer* repo's CI,
+    # so every file lookup and git operation must happen inside a clone of
+    # the target repo, not the producer checkout in CWD.
+    workdir = Path(tempfile.mkdtemp(prefix="dep-pr-"))
+    clone_url = f"{GITEA_API_URL.removesuffix('/api/v1')}/{repo}.git"
+    auth_cfg = f"http.extraHeader=Authorization: token {token}"
+    subprocess.run(  # nosec B603 B607
+        ["git", "-c", auth_cfg, "clone", "--depth", "50", clone_url, str(workdir)],
+        check=True,
+        capture_output=True,
+    )
+
     # Find current pinned version
     old_version = None
     changed_file = None
     for f in [PYPROJECT_PATH, IMAGES_YML_PATH, ROLE_DEFAULTS_PATH]:
-        old_version = find_pinned_version(package, f)
+        old_version = find_pinned_version(package, str(workdir / f))
         if old_version:
             changed_file = f
             break
@@ -184,17 +287,20 @@ def cli(
         else:
             raise click.ClickException(_("Failed to create branch: {error}", error=str(e))) from None
 
-    # Clone, update file, commit, push
-    subprocess.run(["git", "fetch", "origin", f"{branch_name}"], check=False, capture_output=True)  # nosec B603 B607
-    subprocess.run(["git", "checkout", branch_name], check=False, capture_output=True)  # nosec B603 B607
+    # Check out the API-created branch inside the target clone.
+    subprocess.run(["git", "fetch", "origin", f"{branch_name}"], check=False, capture_output=True, cwd=workdir)  # nosec B603 B607
+    subprocess.run(["git", "checkout", branch_name], check=False, capture_output=True, cwd=workdir)  # nosec B603 B607
 
-    if not changed_file or not update_pinned_version(changed_file, package, old_version, new_version):
+    if not changed_file or not update_pinned_version(str(workdir / changed_file), package, old_version, new_version):
         raise click.ClickException(_("Failed to update {file}", file=changed_file))
-
-    subprocess.run(["git", "add", changed_file], check=True)  # nosec B603 B607
+    subprocess.run(["git", "add", changed_file], check=True, cwd=workdir)  # nosec B603 B607
     commit_msg = f"deps: bump {package} from {old_version} to {new_version}"
-    subprocess.run(["git", "commit", "-m", commit_msg], check=True)  # nosec B603 B607
-    subprocess.run(["git", "push", "origin", branch_name], check=True)  # nosec B603 B607
+    subprocess.run(["git", "commit", "-m", commit_msg], check=True, cwd=workdir)  # nosec B603 B607
+    subprocess.run(  # nosec B603 B607
+        ["git", "-c", auth_cfg, "push", "origin", branch_name],
+        check=True,
+        cwd=workdir,
+    )
 
     # Create Vikunja task for tracking
     task_title = f"Bump {package} to {new_version}"

src/devx/ci/nightly_gate.py

@@ -97,10 +97,17 @@ def cli(repo: str, action: str, run_id: str, github_output: bool) -> None:
                 write_github_output("nightly-status", status)
             raise click.ClickException(_("Nightly gate failed — staging deploy blocked."))
         else:
-            click.echo(f"[nightly-gate] Unknown nightly status: {status} — allowing deploy.")
+            click.echo(
+                _(
+                    "[nightly-gate] Unknown nightly status: {status} — blocking deploy (fail closed).",
+                    status=status,
+                ),
+                err=True,
+            )
             if github_output:
-                write_github_output("nightly-gate-passed", "true")
+                write_github_output("nightly-gate-passed", "false")
                 write_github_output("nightly-status", status)
+            raise click.ClickException(_("Unknown nightly status — staging deploy blocked."))
 
     elif action == "set-passed":
         set_nightly_status(client, f"passed:{run_id}" if run_id else "passed")

src/devx/ci/validate_commit_msg.py

@@ -118,7 +118,7 @@ def main(commit_msg_file: str | None, branch: str | None, from_git: bool) -> Non
                 "  Expected: <type>: <description>\n"
                 "  Got:      {subject}\n"
                 "  Allowed types: feat, fix, chore, docs, style, refactor,\n"
-                "                 perf, test, ci, build, revert, BREAKING CHANGE",
+                "                 perf, test, ci, build, deps, revert, BREAKING CHANGE",
                 subject=subject,
             )
         )

src/devx/config.py

@@ -93,4 +93,4 @@ RETRY_BACKOFF_BASE = 2  # seconds: 2, 4, 8
 RETRY_STATUS_CODES = {429, 500, 502, 503, 504}
 
 # Conventional commit regex — used by validate_commit_msg.py
-CONVENTIONAL_RE = re.compile(r"^(feat|fix|chore|docs|style|refactor|perf|test|ci|build|revert)(\(.+\))?: .+")
+CONVENTIONAL_RE = re.compile(r"^(feat|fix|chore|docs|style|refactor|perf|test|ci|build|revert|deps)(\(.+\))?: .+")

src/devx/tools/build_image.py

@@ -269,17 +269,19 @@ def _run_push(cmd: list[str]) -> subprocess.CompletedProcess[str]:
 
     Docker sends push progress/errors to both stdout and stderr depending
     on the error type, so both streams are checked for the 500 status.
+    Uses stderr=STDOUT to merge both streams into stdout, ensuring all
+    output is captured in one place (docker push output behavior varies
+    depending on TTY detection).
     """
     result = subprocess.run(  # nosec B603
         cmd,
-        capture_output=True,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.STDOUT,
         text=True,
         check=False,
     )
-    if result.returncode != 0:
-        combined = f"{result.stderr}\n{result.stdout}"
-        if "500" in combined:
-            raise PushHTTP500Error(combined.strip())
+    if result.returncode != 0 and "500" in (result.stdout or ""):
+        raise PushHTTP500Error(result.stdout.strip())
     return result
 
 
@@ -313,9 +315,17 @@ def push_image(
         click.echo(f"Pushing {ft}...")
 
         @retry(
-            stop=stop_after_attempt(3),
-            wait=wait_exponential(multiplier=5, min=5, max=20),
+            stop=stop_after_attempt(5),
+            wait=wait_exponential(multiplier=10, min=10, max=60),
             retry=retry_if_exception_type(PushHTTP500Error),
+            before_sleep=lambda retry_state: click.echo(
+                _(
+                    "  HTTP 500 from registry, retrying in {wait:.0f}s (attempt {attempt}/5)...",
+                    wait=retry_state.next_action.sleep if retry_state.next_action else 0,
+                    attempt=retry_state.attempt_number + 1,
+                ),
+                err=True,
+            ),
             reraise=True,
         )
         def _attempt(_cmd: list[str] = cmd) -> subprocess.CompletedProcess[str]:
@@ -334,7 +344,7 @@ def push_image(
         if result.returncode == 0:
             click.echo(f"Pushed {ft}")
             continue
-        combined_output = f"{result.stderr}\n{result.stdout}".strip()
+        combined_output = (result.stdout or "").strip()
         # Gitea #31964: push fails because tag already exists.
         # Delete the old manifest and retry once.
         if username and token and "already exists" in combined_output.lower():
@@ -350,14 +360,15 @@ def push_image(
             click.echo(f"  Retrying push {ft}...")
             result = subprocess.run(  # nosec B603
                 cmd,
-                capture_output=True,
+                stdout=subprocess.PIPE,
+                stderr=subprocess.STDOUT,
                 text=True,
                 check=False,
             )
             if result.returncode == 0:
                 click.echo(f"Pushed {ft} (after retry)")
                 continue
-            combined_output = f"{result.stderr}\n{result.stdout}".strip()
+            combined_output = (result.stdout or "").strip()
         click.echo(
             _("Push failed for {tag}: {error}", tag=ft, error=combined_output),
             err=True,

tests/unit/test_api_clients.py

@@ -923,6 +923,12 @@ class TestGiteaClientActions:
         )
 
     def test_get_repo_variable_returns_value(self) -> None:
+        client = GiteaClient("https://git.example.com", "tok", "owner", "repo")
+        client._session.request = MagicMock(return_value=_mock_response({"data": "v0.28.1"}))
+        result = client.get_repo_variable("PRODUCTION_DEPLOY_TAG")
+        assert result == "v0.28.1"
+
+    def test_get_repo_variable_falls_back_to_value(self) -> None:
         client = GiteaClient("https://git.example.com", "tok", "owner", "repo")
         client._session.request = MagicMock(return_value=_mock_response({"value": "v0.28.1"}))
         result = client.get_repo_variable("PRODUCTION_DEPLOY_TAG")

tests/unit/test_auto_merge.py

@@ -468,3 +468,138 @@ def test_main_module_block() -> None:
     exec(compile(source, am.__file__, "exec"), namespace)
     # Verify main is callable
     assert callable(namespace["main"])
+
+
+# -- DEVX-164: Vikunja outage resilience + self-approval diagnostics --
+
+
+class TestVikunjaLookupRetry:
+    """REQ-1: transient Vikunja API failures are retried, not fatal."""
+
+    @patch.dict("os.environ", {"VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.time.sleep")
+    @patch("devx.ci.auto_merge.VikunjaClient")
+    def test_retries_transient_api_error_then_succeeds(self, mock_client_cls: MagicMock, mock_sleep: MagicMock) -> None:
+        """A 404/502 during a Vikunja restart is retried until tasks list."""
+        mock_client = MagicMock()
+        mock_client.list_project_tasks.side_effect = [
+            APIError(404, "Not Found"),
+            APIError(502, "Bad Gateway"),
+            [{"id": 1, "identifier": "DEVX-19", "title": "Add new feature"}],
+        ]
+        mock_client_cls.return_value = mock_client
+        validate_pr_title_matches_vikunja("DEVX-19: Add new feature", "DEVX-19")
+        assert mock_client.list_project_tasks.call_count == 3
+        assert mock_sleep.call_count == 2
+
+    @patch.dict("os.environ", {"VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.time.sleep")
+    @patch("devx.ci.auto_merge.VikunjaClient")
+    def test_retry_exhaustion_propagates_error(self, mock_client_cls: MagicMock, _mock_sleep: MagicMock) -> None:
+        """Persistent outage still fails after the bounded attempt count."""
+        mock_client = MagicMock()
+        mock_client.list_project_tasks.side_effect = APIError(502, "Bad Gateway")
+        mock_client_cls.return_value = mock_client
+        with pytest.raises(APIError, match="Bad Gateway"):
+            validate_pr_title_matches_vikunja("DEVX-19: test", "DEVX-19")
+        assert mock_client.list_project_tasks.call_count == 4
+
+    @patch.dict("os.environ", {"VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.time.sleep")
+    @patch("devx.ci.auto_merge.VikunjaClient")
+    def test_retries_connection_error(self, mock_client_cls: MagicMock, mock_sleep: MagicMock) -> None:
+        """Connection-level failures during restart are also retried."""
+        import requests as req
+
+        mock_client = MagicMock()
+        mock_client.list_project_tasks.side_effect = [
+            req.ConnectionError("refused"),
+            [{"id": 1, "identifier": "DEVX-19", "title": "Found me"}],
+        ]
+        mock_client_cls.return_value = mock_client
+        validate_pr_title_matches_vikunja("DEVX-19: Found me", "DEVX-19")
+        assert mock_sleep.call_count == 1
+
+    @patch.dict("os.environ", {"VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.VikunjaClient")
+    def test_missing_task_still_fails_without_retry_sleep(self, mock_client_cls: MagicMock) -> None:
+        """A healthy Vikunja that simply lacks the task fails as before."""
+        mock_client = MagicMock()
+        mock_client.list_project_tasks.return_value = []
+        mock_client_cls.return_value = mock_client
+        with pytest.raises(click.ClickException, match="Could not find"):
+            validate_pr_title_matches_vikunja("DEVX-99: test", "DEVX-99")
+
+
+class TestMergeApprovalDiagnostics:
+    """REQ-2: merge 405 reports approval state + self-approval remediation."""
+
+    @patch.dict("os.environ", {"CI_GITEA_TOKEN": "tok", "VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.validate_pr_title_matches_vikunja")
+    @patch("devx.ci.auto_merge.GiteaClient")
+    def test_405_without_approvals_shows_self_approval_hint(
+        self, mock_client_cls: MagicMock, _mock_validate: MagicMock, tmp_path, monkeypatch
+    ) -> None:  # type: ignore[no-untyped-def]
+        monkeypatch.chdir(tmp_path)
+        mock_client = MagicMock()
+        mock_client.get_pr_commits.return_value = [
+            {"commit": {"message": "fix: resolve timeout"}},
+        ]
+        mock_client.merge_pr.side_effect = APIError(405, "Does not have enough approvals")
+        mock_client.get_pr_reviews.return_value = []
+        mock_client_cls.return_value = mock_client
+        runner = CliRunner()
+        result = runner.invoke(
+            main,
+            ["DEVX-19-fix-bug", "DEVX-19: Fix timeout", "owner/repo", "7"],
+        )
+        assert result.exit_code != 0
+        assert "Merge failed" in result.output
+        assert "APPROVED" in result.output
+        assert "non-author account" in result.output
+
+    @patch.dict("os.environ", {"CI_GITEA_TOKEN": "tok", "VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.validate_pr_title_matches_vikunja")
+    @patch("devx.ci.auto_merge.GiteaClient")
+    def test_405_with_approvals_omits_hint(
+        self, mock_client_cls: MagicMock, _mock_validate: MagicMock, tmp_path, monkeypatch
+    ) -> None:  # type: ignore[no-untyped-def]
+        monkeypatch.chdir(tmp_path)
+        mock_client = MagicMock()
+        mock_client.get_pr_commits.return_value = [
+            {"commit": {"message": "fix: resolve timeout"}},
+        ]
+        mock_client.merge_pr.side_effect = APIError(405, "Does not have enough approvals")
+        mock_client.get_pr_reviews.return_value = [{"state": "APPROVED", "user": {"login": "kireto"}}]
+        mock_client_cls.return_value = mock_client
+        runner = CliRunner()
+        result = runner.invoke(
+            main,
+            ["DEVX-19-fix-bug", "DEVX-19: Fix timeout", "owner/repo", "7"],
+        )
+        assert result.exit_code != 0
+        assert "Merge failed" in result.output
+        assert "non-author account" not in result.output
+
+    @patch.dict("os.environ", {"CI_GITEA_TOKEN": "tok", "VIKUNJA_TOKEN": "tok"}, clear=True)
+    @patch("devx.ci.auto_merge.validate_pr_title_matches_vikunja")
+    @patch("devx.ci.auto_merge.GiteaClient")
+    def test_405_reviews_fetch_failure_still_raises(
+        self, mock_client_cls: MagicMock, _mock_validate: MagicMock, tmp_path, monkeypatch
+    ) -> None:  # type: ignore[no-untyped-def]
+        """If the reviews lookup itself fails, the merge error still surfaces."""
+        monkeypatch.chdir(tmp_path)
+        mock_client = MagicMock()
+        mock_client.get_pr_commits.return_value = [
+            {"commit": {"message": "fix: resolve timeout"}},
+        ]
+        mock_client.merge_pr.side_effect = APIError(405, "Does not have enough approvals")
+        mock_client.get_pr_reviews.side_effect = APIError(403, "Forbidden")
+        mock_client_cls.return_value = mock_client
+        runner = CliRunner()
+        result = runner.invoke(
+            main,
+            ["DEVX-19-fix-bug", "DEVX-19: Fix timeout", "owner/repo", "7"],
+        )
+        assert result.exit_code != 0
+        assert "Merge failed" in result.output

tests/unit/test_build_image.py

@@ -198,7 +198,7 @@ class TestBuildImage:
 class TestPushImage:
     def test_success(self) -> None:
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest", "1.0"])
-        mock_result = MagicMock(returncode=0, stderr="", stdout="")
+        mock_result = MagicMock(returncode=0, stdout="")
         with patch("devx.tools.build_image.subprocess.run", return_value=mock_result) as mock_run:
             assert push_image(spec, "git.example.com") is True
             assert mock_run.call_count == 2
@@ -206,8 +206,8 @@ class TestPushImage:
     def test_partial_failure(self) -> None:
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest", "1.0"])
         results = [
-            MagicMock(returncode=0, stderr="", stdout=""),
-            MagicMock(returncode=1, stderr="push failed", stdout=""),
+            MagicMock(returncode=0, stdout=""),
+            MagicMock(returncode=1, stdout="push failed"),
         ]
         with patch("devx.tools.build_image.subprocess.run", side_effect=results):
             assert push_image(spec, "git.example.com") is False
@@ -221,7 +221,7 @@ class TestPushImage:
     def test_no_delete_on_success_with_creds(self) -> None:
         """Push-first: no delete needed when push succeeds."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
-        mock_result = MagicMock(returncode=0, stderr="", stdout="")
+        mock_result = MagicMock(returncode=0, stdout="")
         with (
             patch("devx.tools.build_image.subprocess.run", return_value=mock_result),
             patch("devx.tools.build_image.delete_remote_manifest") as mock_del,
@@ -231,7 +231,7 @@ class TestPushImage:
 
     def test_no_delete_without_creds(self) -> None:
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
-        mock_result = MagicMock(returncode=0, stderr="", stdout="")
+        mock_result = MagicMock(returncode=0, stdout="")
         with (
             patch("devx.tools.build_image.subprocess.run", return_value=mock_result),
             patch("devx.tools.build_image.delete_remote_manifest") as mock_del,
@@ -243,8 +243,8 @@ class TestPushImage:
         """Gitea #31964: push fails with 'already exists', delete + retry."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
         results = [
-            MagicMock(returncode=1, stderr="package version already exists", stdout=""),
-            MagicMock(returncode=0, stderr="", stdout=""),
+            MagicMock(returncode=1, stdout="package version already exists"),
+            MagicMock(returncode=0, stdout=""),
         ]
         with (
             patch("devx.tools.build_image.subprocess.run", side_effect=results),
@@ -263,7 +263,7 @@ class TestPushImage:
     def test_no_delete_on_non_already_exists_failure(self) -> None:
         """Push fails for other reasons (non-500) — old manifest preserved."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
-        mock_result = MagicMock(returncode=1, stderr="denied: requested access to the resource is denied", stdout="")
+        mock_result = MagicMock(returncode=1, stdout="denied: requested access to the resource is denied")
         with (
             patch("devx.tools.build_image.subprocess.run", return_value=mock_result),
             patch("devx.tools.build_image.delete_remote_manifest") as mock_del,
@@ -275,8 +275,8 @@ class TestPushImage:
         """Gitea #31964 retry also fails — both pushes fail."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
         results = [
-            MagicMock(returncode=1, stderr="package version already exists", stdout=""),
-            MagicMock(returncode=1, stderr="push failed again", stdout=""),
+            MagicMock(returncode=1, stdout="package version already exists"),
+            MagicMock(returncode=1, stdout="push failed again"),
         ]
         with (
             patch("devx.tools.build_image.subprocess.run", side_effect=results),
@@ -288,8 +288,8 @@ class TestPushImage:
         """HTTP 500 from registry race condition — retry succeeds."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
         results = [
-            MagicMock(returncode=1, stderr="", stdout="received unexpected HTTP status: 500 Internal Server Error"),
-            MagicMock(returncode=0, stderr="", stdout=""),
+            MagicMock(returncode=1, stdout="received unexpected HTTP status: 500 Internal Server Error"),
+            MagicMock(returncode=0, stdout=""),
         ]
         with (
             patch("devx.tools.build_image.subprocess.run", side_effect=results),
@@ -302,9 +302,7 @@ class TestPushImage:
     def test_http_500_retries_all_fail(self) -> None:
         """HTTP 500 retries exhausted — push fails, no delete attempted."""
         spec = ImageSpec(name="ci-base", dockerfile="Dockerfile", tags=["latest"])
-        mock_result = MagicMock(
-            returncode=1, stderr="", stdout="received unexpected HTTP status: 500 Internal Server Error"
-        )
+        mock_result = MagicMock(returncode=1, stdout="received unexpected HTTP status: 500 Internal Server Error")
         with (
             patch("devx.tools.build_image.subprocess.run", return_value=mock_result),
             patch("devx.tools.build_image.delete_remote_manifest") as mock_del,
@@ -313,31 +311,20 @@ class TestPushImage:
             assert push_image(spec, "git.example.com", username="user", token="tok") is False
             mock_del.assert_not_called()
 
-    def test_run_push_raises_on_500_stderr(self) -> None:
-        """_run_push raises PushHTTP500Error when stderr contains 500."""
+    def test_run_push_raises_on_500(self) -> None:
+        """_run_push raises PushHTTP500Error when stdout contains 500."""
         from devx.tools.build_image import _run_push
 
-        mock_result = MagicMock(returncode=1, stderr="HTTP 500 Internal Server Error", stdout="")
-        with patch("devx.tools.build_image.subprocess.run", return_value=mock_result):
-            with pytest.raises(PushHTTP500Error, match="HTTP 500"):
-                _run_push(["docker", "push", "img:latest"])
-
-    def test_run_push_raises_on_500_stdout(self) -> None:
-        """_run_push raises PushHTTP500Error when stdout contains 500 (docker sends to stdout)."""
-        from devx.tools.build_image import _run_push
-
-        mock_result = MagicMock(
-            returncode=1, stderr="", stdout="received unexpected HTTP status: 500 Internal Server Error"
-        )
+        mock_result = MagicMock(returncode=1, stdout="received unexpected HTTP status: 500 Internal Server Error")
         with patch("devx.tools.build_image.subprocess.run", return_value=mock_result):
             with pytest.raises(PushHTTP500Error, match="500"):
                 _run_push(["docker", "push", "img:latest"])
 
     def test_run_push_no_raise_on_non_500(self) -> None:
-        """_run_push returns result when stderr has no 500."""
+        """_run_push returns result when stdout has no 500."""
         from devx.tools.build_image import _run_push
 
-        mock_result = MagicMock(returncode=1, stderr="denied: access denied", stdout="")
+        mock_result = MagicMock(returncode=1, stdout="denied: access denied")
         with patch("devx.tools.build_image.subprocess.run", return_value=mock_result):
             result = _run_push(["docker", "push", "img:latest"])
             assert result.returncode == 1
@@ -346,7 +333,7 @@ class TestPushImage:
         """_run_push returns result on success."""
         from devx.tools.build_image import _run_push
 
-        mock_result = MagicMock(returncode=0, stderr="", stdout="")
+        mock_result = MagicMock(returncode=0, stdout="")
         with patch("devx.tools.build_image.subprocess.run", return_value=mock_result):
             result = _run_push(["docker", "push", "img:latest"])
             assert result.returncode == 0

tests/unit/test_check_pr_size.py

@@ -26,6 +26,12 @@ class TestIsExcluded:
     def test_excludes_readme(self) -> None:
         assert is_excluded("README.md", ["README.md"])
 
+    def test_excludes_plans_glob(self) -> None:
+        assert is_excluded("docs/plans/sso-bridge-full-extraction.md", ["docs/plans/*"])
+
+    def test_does_not_exclude_specs(self) -> None:
+        assert not is_excluded("docs/specs/DEVX-165.md", ["docs/plans/*"])
+
 
 class TestCheckSize:
     def test_under_limits_passes(self) -> None:

tests/unit/test_config.py

@@ -38,6 +38,7 @@ class TestConfigConstants:
     def test_conventional_re(self) -> None:
         assert CONVENTIONAL_RE.match("feat: add feature")
         assert CONVENTIONAL_RE.match("fix(scope): bug fix")
+        assert CONVENTIONAL_RE.match("deps: bump devx from v0.50.2 to v0.51.0")
         assert not CONVENTIONAL_RE.match("random message")
         assert not CONVENTIONAL_RE.match("feat:")
         assert not CONVENTIONAL_RE.match("BREAKING CHANGE: something")

tests/unit/test_create_dependency_pr.py

@@ -4,6 +4,7 @@ from pathlib import Path
 from unittest.mock import MagicMock, patch
 
 import click
+import pytest
 from click.testing import CliRunner
 
 from devx.ci.create_dependency_pr import (
@@ -15,6 +16,13 @@ from devx.ci.create_dependency_pr import (
 )
 
 
+@pytest.fixture(autouse=True)
+def _mock_subprocess():
+    """Mock subprocess so CLI tests never run a real git clone."""
+    with patch("devx.ci.create_dependency_pr.subprocess.run") as m:
+        yield m
+
+
 class TestFindPinnedVersion:
     def test_finds_pip_git_pin(self, tmp_path: Path) -> None:
         content = "grm @ git+https://git.example.com/repo.git@v0.5.1"
@@ -180,3 +188,164 @@ class TestCreateVikunjaTask:
             mock_client.create_task.return_value = {"identifier": "OBL-INFRA-999"}
             result = create_vikunja_task("Test", "desc")
             assert result == "OBL-INFRA-999"
+
+
+class TestResolveContainerDigest:
+    """REQ-1: pre-PR artifact verification via the packages API."""
+
+    def test_returns_digest_from_manifest_blob(self) -> None:
+        from devx.ci.create_dependency_pr import resolve_container_digest
+
+        mock_resp = MagicMock()
+        mock_resp.raise_for_status = MagicMock()
+        mock_resp.json.return_value = [
+            {"name": "sha256_layer", "sha256": "abc"},
+            {"name": "manifest.json", "sha256": "deadbeef"},
+        ]
+        with patch("devx.ci.create_dependency_pr.requests.get", return_value=mock_resp):
+            digest = resolve_container_digest(
+                "https://git.example.com/api/v1", "oblachno", "sso-bridge", "0.9.1", "tok"
+            )
+        assert digest == "sha256:deadbeef"
+
+    def test_raises_when_version_missing(self) -> None:
+        import requests
+
+        from devx.ci.create_dependency_pr import resolve_container_digest
+
+        mock_resp = MagicMock()
+        http_err = requests.HTTPError("404")
+        http_err.response = MagicMock(status_code=404)
+        mock_resp.raise_for_status.side_effect = http_err
+        with patch("devx.ci.create_dependency_pr.requests.get", return_value=mock_resp):
+            with pytest.raises(click.ClickException, match="unpublished artifact"):
+                resolve_container_digest("https://git.example.com/api/v1", "oblachno", "sso-bridge", "9.9.9", "tok")
+
+    def test_raises_when_no_manifest_blob(self) -> None:
+        from devx.ci.create_dependency_pr import resolve_container_digest
+
+        mock_resp = MagicMock()
+        mock_resp.raise_for_status = MagicMock()
+        mock_resp.json.return_value = [{"name": "sha256_layer", "sha256": "abc"}]
+        with patch("devx.ci.create_dependency_pr.requests.get", return_value=mock_resp):
+            with pytest.raises(click.ClickException, match="no manifest blob"):
+                resolve_container_digest("https://git.example.com/api/v1", "oblachno", "sso-bridge", "0.9.1", "tok")
+
+    def test_raises_on_connection_error(self) -> None:
+        import requests
+
+        from devx.ci.create_dependency_pr import resolve_container_digest
+
+        with patch(
+            "devx.ci.create_dependency_pr.requests.get",
+            side_effect=requests.ConnectionError("refused"),
+        ):
+            with pytest.raises(click.ClickException, match="Registry lookup failed"):
+                resolve_container_digest("https://git.example.com/api/v1", "oblachno", "sso-bridge", "0.9.1", "tok")
+
+
+class TestCliVerifyContainer:
+    """REQ-1: artifact verification gates the dependency PR."""
+
+    @patch("devx.ci.create_dependency_pr.resolve_container_digest")
+    @patch("devx.ci.create_dependency_pr.find_pinned_version")
+    @patch("devx.ci.create_dependency_pr.get_ci_token")
+    def test_verify_container_runs_before_lookup(
+        self, mock_token: MagicMock, mock_find: MagicMock, mock_digest: MagicMock
+    ) -> None:
+        """Verification failure aborts before the version lookup/PR steps."""
+        mock_token.return_value = "fake-token"
+        mock_digest.side_effect = click.ClickException("unpublished artifact")
+        runner = CliRunner()
+        result = runner.invoke(
+            cli,
+            [
+                "--package",
+                "sso_bridge",
+                "--new-version",
+                "0.9.1",
+                "--source-repo",
+                "oblachno/sso-bridge",
+                "--verify-container",
+                "oblachno/sso-bridge",
+            ],
+        )
+        assert result.exit_code != 0
+        mock_find.assert_not_called()
+
+    @patch("devx.ci.create_dependency_pr.resolve_container_digest")
+    @patch("devx.ci.create_dependency_pr.find_pinned_version")
+    @patch("devx.ci.create_dependency_pr.get_ci_token")
+    def test_verify_container_resolves_digest(
+        self, mock_token: MagicMock, mock_find: MagicMock, mock_digest: MagicMock
+    ) -> None:
+        mock_token.return_value = "fake-token"
+        mock_find.return_value = "0.9.1"
+        mock_digest.return_value = "sha256:abc"
+        runner = CliRunner()
+        result = runner.invoke(
+            cli,
+            [
+                "--package",
+                "sso_bridge",
+                "--new-version",
+                "0.9.1",
+                "--source-repo",
+                "oblachno/sso-bridge",
+                "--verify-container",
+                "oblachno/sso-bridge",
+            ],
+        )
+        assert result.exit_code == 0
+        mock_digest.assert_called_once()
+        args = mock_digest.call_args[0]
+        assert args[1:4] == ("oblachno", "sso-bridge", "0.9.1")
+
+    @patch("devx.ci.create_dependency_pr.get_ci_token")
+    def test_verify_container_invalid_format(self, mock_token: MagicMock) -> None:
+        mock_token.return_value = "fake-token"
+        runner = CliRunner()
+        result = runner.invoke(
+            cli,
+            [
+                "--package",
+                "sso_bridge",
+                "--new-version",
+                "0.9.1",
+                "--source-repo",
+                "oblachno/sso-bridge",
+                "--verify-container",
+                "no-slash",
+            ],
+        )
+        assert result.exit_code != 0
+
+    @patch("devx.ci.create_dependency_pr.resolve_container_digest")
+    @patch("devx.ci.create_dependency_pr.find_pinned_version")
+    @patch("devx.ci.create_dependency_pr.get_ci_token")
+    def test_container_tag_overrides_new_version(
+        self, mock_token: MagicMock, mock_find: MagicMock, mock_digest: MagicMock
+    ) -> None:
+        """--container-tag selects the image tag when it differs from version."""
+        mock_token.return_value = "fake-token"
+        mock_find.return_value = "0.9.1"
+        mock_digest.return_value = "sha256:abc"
+        runner = CliRunner()
+        result = runner.invoke(
+            cli,
+            [
+                "--package",
+                "sso_bridge",
+                "--new-version",
+                "0.9.1",
+                "--source-repo",
+                "oblachno/sso-bridge",
+                "--verify-container",
+                "oblachno/sso-bridge",
+                "--container-tag",
+                "0.2.4",
+            ],
+        )
+        assert result.exit_code == 0
+        args = mock_digest.call_args[0]
+        assert args[1:4] == ("oblachno", "sso-bridge", "0.2.4")

tests/unit/test_nightly_gate.py

@@ -71,6 +71,17 @@ class TestCli:
         assert result.exit_code != 0
         assert "blocked" in result.output.lower()
 
+    @patch("devx.ci.nightly_gate.GiteaClient")
+    @patch("devx.ci.nightly_gate.get_ci_token")
+    def test_check_unknown_status_blocks_deploy(self, mock_token: MagicMock, mock_client_cls: MagicMock) -> None:
+        mock_token.return_value = "fake-token"
+        mock_client = mock_client_cls.return_value
+        mock_client.get_repo_variable.return_value = "garbage-value"
+        runner = CliRunner()
+        result = runner.invoke(cli, ["--repo", "oblachno/infra", "--action", "check"])
+        assert result.exit_code != 0
+        assert "fail closed" in result.output.lower()
+
     @patch("devx.ci.nightly_gate.GiteaClient")
     @patch("devx.ci.nightly_gate.get_ci_token")
     def test_set_passed(self, mock_token: MagicMock, mock_client_cls: MagicMock) -> None:

tests/unit/test_skills_validation.py

@@ -0,0 +1,121 @@
+"""Pytest tests for Devin skill validation.
+
+Validates that all skills in .devin/skills/ are well-formed: H1 title,
+"when to invoke" section, prerequisites when commands are referenced,
+make-target references that exist, and file references that exist.
+
+Run with: make pytest TEST=tests/test_skills_validation.py
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+import pytest
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+
+# Sections required for every skill
+REQUIRED_SECTIONS = ["when to invoke"]
+
+# Sections required only for skills that reference commands/tools
+COMMAND_REQUIRED_SECTIONS = ["prerequisites"]
+
+# Markers indicating a skill references commands/tools
+COMMAND_MARKERS = ("`make ", "```bash", "```sh", "curl ", "python ", "python3 ", "ssh ")
+
+EXPECTED_SKILLS = [
+    "dependency-graph",
+    "deployment-coordination",
+    "devx-workflow",
+    "pr-review",
+    "skill-creation",
+    "spec-driven-development",
+    "testing-and-debugging",
+    "vikunja-tasks",
+]
+
+
+def _find_skills() -> dict[str, Path]:
+    skills_dir = REPO_ROOT / ".devin" / "skills"
+    assert skills_dir.exists(), ".devin/skills/ directory not found"
+    return {d.name: d / "SKILL.md" for d in skills_dir.iterdir() if d.is_dir() and (d / "SKILL.md").exists()}
+
+
+# Skills shared with other repos — file-path references are only checked
+# in the owning repo (infra), where the referenced files live.
+SHARED_SKILLS = {"cross-repo-sync", "branch-hygiene", "dependency-graph", "skill-creation"}
+
+
+def _make_targets() -> set[str]:
+    """Collect make targets from Makefile plus included devx .mak files."""
+    targets: set[str] = set()
+    makefile = REPO_ROOT / "Makefile"
+    if makefile.exists():
+        targets.update(re.findall(r"^([a-zA-Z][a-zA-Z0-9_-]*):", makefile.read_text(), re.MULTILINE))
+    for mak in REPO_ROOT.glob(".venv/lib/python*/site-packages/devx/make/*.mak"):
+        targets.update(re.findall(r"^([a-zA-Z][a-zA-Z0-9_-]*):", mak.read_text(), re.MULTILINE))
+    # devx repo: devx.mak lives in the package source (editable install)
+    for mak in REPO_ROOT.glob("src/devx/make/*.mak"):
+        targets.update(re.findall(r"^([a-zA-Z][a-zA-Z0-9_-]*):", mak.read_text(), re.MULTILINE))
+    return targets
+
+
+def _validate_skill(skill_name: str, skill_path: Path, make_targets: set[str]) -> list[str]:
+    """Validate a single skill file. Returns list of error messages."""
+    errors: list[str] = []
+    content = skill_path.read_text()
+
+    if not re.search(r"^# ", content, re.MULTILINE):
+        errors.append(f"{skill_name}: missing H1 title")
+
+    lower = content.lower()
+    for section in REQUIRED_SECTIONS:
+        if f"## {section}" not in lower:
+            errors.append(f"{skill_name}: missing '## {section.title()}' section")
+
+    references_commands = any(marker in content for marker in COMMAND_MARKERS)
+    if references_commands:
+        for section in COMMAND_REQUIRED_SECTIONS:
+            if f"## {section}" not in lower:
+                errors.append(
+                    f"{skill_name}: missing '## {section.title()}' section "
+                    "(required because skill references commands/tools)"
+                )
+
+    for target in re.findall(r"`make ([a-zA-Z][a-zA-Z0-9_-]*)`", content):
+        if target not in make_targets:
+            errors.append(f"{skill_name}: references `make {target}` but target does not exist")
+
+    # File-path checks: skip shared skills (checked in infra) and
+    # placeholder paths containing <...> templates.
+    if skill_name not in SHARED_SKILLS:
+        for match in re.findall(r"`((?:scripts|src|ansible|docs|tests|environments)/[^`\s]+)`", content):
+            if "<" in match:
+                continue
+            if not (REPO_ROOT / match).exists():
+                errors.append(f"{skill_name}: references `{match}` but file does not exist")
+
+    return errors
+
+
+@pytest.mark.parametrize("skill_name", EXPECTED_SKILLS)
+def test_skill_exists(skill_name: str) -> None:
+    """Each expected skill must have a SKILL.md."""
+    skill = REPO_ROOT / ".devin" / "skills" / skill_name / "SKILL.md"
+    assert skill.exists(), f"{skill_name}/SKILL.md not found"
+
+
+def test_minimum_skill_count() -> None:
+    """The repo should carry a working set of skills, not a stub."""
+    assert len(_find_skills()) >= 7, "expected >=10 skills"
+
+
+def test_all_skills_validate() -> None:
+    """All skills must pass structure/reference validation."""
+    make_targets = _make_targets()
+    errors: list[str] = []
+    for skill_name, skill_path in _find_skills().items():
+        errors.extend(_validate_skill(skill_name, skill_path, make_targets))
+    assert not errors, "Skill validation failed:\n" + "\n".join(f"  - {e}" for e in errors)

tests/unit/test_validate_commit_msg.py

@@ -30,6 +30,7 @@ class TestHelpers:
         assert CONVENTIONAL_RE.match("test: add tests")
         assert CONVENTIONAL_RE.match("ci: update workflow")
         assert CONVENTIONAL_RE.match("build: update deps")
+        assert CONVENTIONAL_RE.match("deps: bump devx from v0.50.2 to v0.51.0")
         assert CONVENTIONAL_RE.match("revert: undo change")
 
     def test_conventional_re_allows_scope(self) -> None: