@htekdev/actions-debugger 1.0.83 → 1.0.84

package/errors/concurrency-timing/event-number-empty-on-push-collapses-concurrency.yml

@@ -0,0 +1,90 @@
+id: concurrency-timing-042
+title: "github.event.number empty on push events collapses concurrency group across all branches"
+category: concurrency-timing
+severity: silent-failure
+tags:
+  - concurrency
+  - github-event-number
+  - push-event
+  - cancel-in-progress
+  - cross-branch
+  - pull-request
+patterns:
+  - regex: 'github\.event\.number\b'
+    flags: 'i'
+error_messages:
+  - "Run was cancelled due to concurrency group collision between push events on different branches"
+  - "Unexpected cancellation of push workflow by unrelated branch push"
+root_cause: |
+  `github.event.number` contains the pull request or issue number for events that have one
+  (e.g., `pull_request`, `pull_request_review`, `issue_comment`). On `push` events,
+  `github.event.number` is undefined and evaluates to an empty string in expressions.
+
+  When a workflow is triggered by both `push` and `pull_request` events, and the
+  concurrency group key uses `github.event.number`, all push-triggered runs share a
+  single concurrency group key (the suffix becomes empty):
+
+    group: "${{ github.workflow }}-${{ github.event.number }}"
+    # PR run:   "CI-123"   (unique per PR number)
+    # Push run: "CI-"      (ALL branch pushes share this single group!)
+
+  With `cancel-in-progress: true`, pushing to any branch cancels all other in-progress
+  push runs — even if they are on completely different branches. Developers intend to
+  serialize CI per PR but instead unintentionally cancel unrelated branch pushes.
+fix: |
+  Use the `||` operator to fall back to `github.ref` when `github.event.number` is empty.
+  `github.ref` is always populated (e.g., `refs/heads/main`, `refs/pull/123/merge`) and
+  produces a unique key per branch or PR.
+
+  Alternatively, use `github.event.number || github.sha` if you want push runs to each
+  be independent (no cancellation between pushes) while still serializing per PR.
+fix_code:
+  - language: yaml
+    label: "Wrong — all push runs share the same concurrency group key"
+    code: |
+      on:
+        push:
+          branches: [main, 'feature/**']
+        pull_request:
+
+      concurrency:
+        group: ${{ github.workflow }}-${{ github.event.number }}
+        # On push: group = "CI-"  (empty number → all pushes share one group)
+        # On PR:   group = "CI-123" (unique per PR)
+        cancel-in-progress: true
+  - language: yaml
+    label: "Correct — fall back to github.ref so push runs are keyed per branch"
+    code: |
+      on:
+        push:
+          branches: [main, 'feature/**']
+        pull_request:
+
+      concurrency:
+        group: ${{ github.workflow }}-${{ github.event.number || github.ref }}
+        # On push: group = "CI-refs/heads/main" (unique per branch)
+        # On PR:   group = "CI-123" (unique per PR number)
+        cancel-in-progress: true
+  - language: yaml
+    label: "Alternative — use github.ref for both triggers (no per-PR serialization)"
+    code: |
+      on:
+        push:
+          branches: [main, 'feature/**']
+        pull_request:
+
+      concurrency:
+        group: ${{ github.workflow }}-${{ github.ref }}
+        # On push: group = "CI-refs/heads/main"
+        # On PR:   group = "CI-refs/pull/123/merge"
+        cancel-in-progress: true
+prevention:
+  - "Always test concurrency group keys against all event types in the on: block"
+  - "Use github.event.number || github.ref as a safe pattern for mixed push/PR workflows"
+  - "Check that github.event.number is not undefined by printing it in a debug step when first setting up concurrency"
+  - "Prefer github.ref or github.ref_name as the base concurrency discriminator for branch-level serialization"
+docs:
+  - url: "https://docs.github.com/en/actions/writing-workflows/workflow-syntax-for-github-actions#jobsjob_idconcurrency"
+    label: "Workflow syntax: concurrency"
+  - url: "https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/accessing-contextual-information-about-workflow-runs#github-context"
+    label: "GitHub context — github.event.number availability by event type"

package/errors/known-unsolved/workflow-dispatch-api-no-run-id-in-response.yml

@@ -0,0 +1,101 @@
+id: known-unsolved-050
+title: "workflow_dispatch REST API dispatch returns 204 No Content — triggered run_id unavailable"
+category: known-unsolved
+severity: limitation
+tags:
+  - workflow_dispatch
+  - rest-api
+  - run-id
+  - polling
+  - automation
+  - cd-pipeline
+patterns:
+  - regex: 'POST.*dispatches.*204|dispatches.*No Content'
+    flags: 'i'
+error_messages:
+  - "204 No Content"
+  - "Unable to correlate triggered workflow run"
+  - "no run_id in dispatch response"
+root_cause: |
+  The GitHub Actions REST API endpoint
+  `POST /repos/{owner}/{repo}/actions/workflows/{workflow_id}/dispatches`
+  always returns HTTP 204 No Content on success — there is no run_id or any identifier
+  in the response body. GitHub creates the workflow run asynchronously after the API call
+  returns, so the run may not exist in the system for several seconds. There is no built-in
+  way to correlate a `workflow_dispatch` API call to the resulting run_id.
+
+  This is a persistent limitation reported since 2020 on the GitHub feedback forum with
+  thousands of upvotes. Developers building CD pipelines that need to trigger a workflow
+  and then monitor or gate on its result are forced to use fragile polling heuristics.
+fix: |
+  No direct fix — this is a GitHub platform limitation. Use one of these workarounds:
+
+  1. **Inject a correlation ID via inputs**: Pass a unique value (e.g., UUID) as a
+     `workflow_dispatch` input. After dispatch, poll `GET /repos/.../actions/runs?event=workflow_dispatch&branch=<branch>`
+     and match runs whose `inputs` contain the correlation ID.
+
+  2. **Use workflow_run event callback**: Design the triggered workflow to complete and
+     then fire a `workflow_run` event that your orchestrating workflow listens for.
+
+  3. **Switch to repository_dispatch**: `repository_dispatch` with a unique `client_payload`
+     field lets you include a correlation ID. The triggered workflow can write it to an artifact
+     or output for later retrieval.
+
+  4. **Third-party action**: Use `peter-evans/workflow-dispatch@v3` which implements polling
+     heuristics to approximate the run_id after dispatch.
+fix_code:
+  - language: yaml
+    label: "Inject correlation ID via workflow_dispatch input — then poll to find run"
+    code: |
+      # Triggering workflow or script passes a unique correlation ID:
+      # POST /repos/{owner}/{repo}/actions/workflows/deploy.yml/dispatches
+      # { "ref": "main", "inputs": { "correlation_id": "abc-123-uuid" } }
+
+      # Triggered workflow (deploy.yml) writes ID to output artifact:
+      on:
+        workflow_dispatch:
+          inputs:
+            correlation_id:
+              description: 'Unique ID for run correlation'
+              required: false
+              default: ''
+
+      jobs:
+        deploy:
+          runs-on: ubuntu-latest
+          steps:
+            - run: echo "run_id=${{ github.run_id }}" >> correlation.txt
+            - uses: actions/upload-artifact@v4
+              with:
+                name: correlation-${{ inputs.correlation_id }}
+                path: correlation.txt
+  - language: yaml
+    label: "Use repository_dispatch with client_payload for reliable correlation"
+    code: |
+      # Trigger with unique client_payload via API:
+      # POST /repos/{owner}/{repo}/dispatches
+      # { "event_type": "deploy", "client_payload": { "job_id": "build-789" } }
+
+      on:
+        repository_dispatch:
+          types: [deploy]
+
+      jobs:
+        deploy:
+          runs-on: ubuntu-latest
+          steps:
+            - run: |
+                echo "Triggered by job: ${{ github.event.client_payload.job_id }}"
+                echo "This run ID: ${{ github.run_id }}"
+prevention:
+  - "Design CD pipelines using workflow_run event callbacks instead of polling after dispatch"
+  - "Include a unique correlation_id input in all workflow_dispatch-triggered workflows"
+  - "Track the GitHub public roadmap — run_id in dispatch response is a long-requested feature"
+  - "Consider repository_dispatch for machine-to-machine triggers needing reliable correlation"
+docs:
+  - url: "https://docs.github.com/en/rest/actions/workflows?apiVersion=2022-11-28#create-a-workflow-dispatch-event"
+    label: "REST API: Create a workflow dispatch event (returns 204 No Content)"
+  - url: "https://docs.github.com/en/actions/writing-workflows/choosing-when-your-workflow-runs/events-that-trigger-workflows#workflow_run"
+    label: "workflow_run event — callback pattern for monitoring triggered workflows"
+  - url: "https://github.com/peter-evans/workflow-dispatch"
+    label: "peter-evans/workflow-dispatch — community action with polling heuristics"

package/errors/silent-failures/skipped-step-job-output-silently-empty.yml

@@ -0,0 +1,119 @@
+id: silent-failures-078
+title: "Skipped step leaves job-level outputs: silently empty — downstream jobs receive empty string"
+category: silent-failures
+severity: silent-failure
+tags:
+  - job-outputs
+  - skipped-step
+  - outputs
+  - if-condition
+  - conditional
+  - downstream-jobs
+patterns:
+  - regex: 'steps\.[a-z_][a-z0-9_-]*\.outputs\.[a-z_]'
+    flags: 'i'
+error_messages:
+  - "needs.<job>.outputs.<key> is empty string when producing step was skipped"
+  - "Job output silently empty — step that writes output was conditionally skipped"
+root_cause: |
+  A job's `outputs:` block maps keys to values using `${{ steps.<id>.outputs.<key> }}`
+  expressions. These expressions are evaluated at the end of the job, after all steps have
+  run. If the step referenced in the output mapping is skipped (due to an `if:` condition
+  evaluating to false), the step never runs and never writes to `$GITHUB_OUTPUT`. The
+  output expression evaluates to an empty string with no warning.
+
+  Downstream jobs that depend on this output via `needs.<job>.outputs.<key>` receive an
+  empty string and continue running without error. This silent empty propagation causes
+  incorrect behavior: deploy jobs receiving an empty version string, notification jobs
+  receiving an empty status, or matrix jobs receiving an empty configuration list.
+
+  This is distinct from the `skipped-job-outputs-empty-string` limitation (where the
+  entire job is skipped) — here the job itself runs successfully, but one or more of its
+  steps are conditionally skipped.
+fix: |
+  Ensure the step that writes the output always runs, or provide a fallback value.
+  Use `if: always()` on the step if it should run regardless of prior step results.
+  Use a separate step to set a default/fallback output for the skipped case.
+  Alternatively, restructure the workflow so that the output-producing step is never
+  guarded by a condition that might prevent it from running.
+fix_code:
+  - language: yaml
+    label: "Wrong — output step conditionally skipped, downstream sees empty string"
+    code: |
+      jobs:
+        build:
+          runs-on: ubuntu-latest
+          outputs:
+            version: ${{ steps.get-version.outputs.version }}
+          steps:
+            - uses: actions/checkout@v4
+
+            - name: Get version (only on tags)
+              id: get-version
+              if: startsWith(github.ref, 'refs/tags/')   # skipped on branch pushes!
+              run: echo "version=${GITHUB_REF_NAME}" >> $GITHUB_OUTPUT
+
+        deploy:
+          needs: build
+          runs-on: ubuntu-latest
+          steps:
+            - run: |
+                echo "Deploying version: ${{ needs.build.outputs.version }}"
+                # On branch push: prints "Deploying version: " (empty — step was skipped)
+  - language: yaml
+    label: "Correct — provide a fallback step to always set the output"
+    code: |
+      jobs:
+        build:
+          runs-on: ubuntu-latest
+          outputs:
+            version: ${{ steps.get-version.outputs.version }}
+          steps:
+            - uses: actions/checkout@v4
+
+            - name: Get version from tag
+              id: get-version
+              run: |
+                if [[ "${GITHUB_REF}" == refs/tags/* ]]; then
+                  echo "version=${GITHUB_REF_NAME}" >> $GITHUB_OUTPUT
+                else
+                  echo "version=dev-${GITHUB_SHA::8}" >> $GITHUB_OUTPUT
+                fi
+              # Always runs — handles both tag and branch pushes
+
+        deploy:
+          needs: build
+          runs-on: ubuntu-latest
+          steps:
+            - run: echo "Deploying version ${{ needs.build.outputs.version }}"
+  - language: yaml
+    label: "Alternative — gate deploy job to only run when version is set"
+    code: |
+      jobs:
+        build:
+          runs-on: ubuntu-latest
+          outputs:
+            version: ${{ steps.get-version.outputs.version }}
+          steps:
+            - uses: actions/checkout@v4
+            - name: Get version (only on tags)
+              id: get-version
+              if: startsWith(github.ref, 'refs/tags/')
+              run: echo "version=${GITHUB_REF_NAME}" >> $GITHUB_OUTPUT
+
+        deploy:
+          needs: build
+          if: needs.build.outputs.version != ''   # only run when version was set
+          runs-on: ubuntu-latest
+          steps:
+            - run: echo "Deploying ${{ needs.build.outputs.version }}"
+prevention:
+  - "Audit all job outputs: blocks — for each step reference, verify it always runs"
+  - "Avoid conditional if: guards on steps that produce job-level outputs unless a fallback step provides the same output key"
+  - "Add an explicit check step at the end of each job that validates required outputs are non-empty"
+  - "Use actionlint — it warns when output-producing steps have if: conditions that may skip them"
+docs:
+  - url: "https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/passing-information-between-jobs"
+    label: "Passing information between jobs — job outputs"
+  - url: "https://docs.github.com/en/actions/writing-workflows/choosing-when-your-workflow-runs/events-that-trigger-workflows"
+    label: "Events that trigger workflows — job outputs with conditional steps"

package/errors/triggers/schedule-paths-filter-silently-ignored.yml

@@ -0,0 +1,111 @@
+id: triggers-058
+title: "on.schedule ignores paths: filter — scheduled workflows always run"
+category: triggers
+severity: silent-failure
+tags:
+  - schedule
+  - paths
+  - cron
+  - trigger
+  - filter
+  - paths-ignore
+patterns:
+  - regex: 'schedule.*paths:|paths:.*cron:'
+    flags: 'si'
+error_messages:
+  - "Scheduled workflow runs even when no matching files changed"
+  - "paths: filter has no effect on scheduled events"
+root_cause: |
+  The `paths:` and `paths-ignore:` filters are only evaluated for `push` and `pull_request`
+  events, where GitHub can compare the changed files in the triggering commit against the
+  filter patterns. `on.schedule` events are time-based and have no associated commit or
+  changeset — there are no paths to compare. GitHub silently ignores any `paths:` or
+  `paths-ignore:` configuration placed under a `schedule:` trigger.
+
+  This commonly happens when a developer copies a `push:` trigger block that includes
+  `paths:` filtering and applies it to a `schedule:` block expecting the same filtering
+  behavior. The workflow runs on every scheduled interval regardless of what files have
+  changed. The `paths:` keys are not flagged as invalid — they are simply ignored without
+  any warning in the workflow or run logs.
+fix: |
+  Remove `paths:` and `paths-ignore:` from `on.schedule:` blocks — they have no effect.
+
+  If you need the scheduled workflow to skip execution when certain conditions are not
+  met, implement gate logic inside the workflow:
+
+  - Use an early job that checks file modification times or an external flag and sets
+    an output. Downstream jobs use `if: needs.check.outputs.should_run == 'true'`.
+  - Use the `dorny/paths-filter` action to detect file changes since a known reference
+    (e.g., the last successful run's commit SHA stored in a cache or artifact).
+  - If the schedule purpose is unrelated to file changes (e.g., dependency audits,
+    health checks), no conditional gate is needed — let it run on schedule.
+fix_code:
+  - language: yaml
+    label: "Wrong — paths: under schedule: is silently ignored"
+    code: |
+      on:
+        schedule:
+          - cron: '0 2 * * *'
+        paths:               # silently ignored — schedule always runs
+          - 'src/**'
+          - 'package.json'
+        push:
+          branches: [main]
+          paths:             # this paths: applies only to the push: trigger
+            - 'src/**'
+            - 'package.json'
+  - language: yaml
+    label: "Correct — paths: filter on push only; schedule unconditional or gated in-workflow"
+    code: |
+      on:
+        schedule:
+          - cron: '0 2 * * *'    # no paths: here — schedule always runs
+        push:
+          branches: [main]
+          paths:                  # paths: filter applies to push trigger only
+            - 'src/**'
+            - 'package.json'
+
+      jobs:
+        build:
+          runs-on: ubuntu-latest
+          steps:
+            - uses: actions/checkout@v4
+            - run: npm run build
+
+      # To conditionally skip a scheduled run, use an in-workflow gate:
+      # jobs:
+      #   check-changes:
+      #     runs-on: ubuntu-latest
+      #     outputs:
+      #       changed: ${{ steps.filter.outputs.src }}
+      #     steps:
+      #       - uses: actions/checkout@v4
+      #         with:
+      #           fetch-depth: 2
+      #       - uses: dorny/paths-filter@v3
+      #         id: filter
+      #         with:
+      #           filters: |
+      #             src:
+      #               - 'src/**'
+      #
+      #   build:
+      #     needs: check-changes
+      #     if: needs.check-changes.outputs.changed == 'true' || github.event_name == 'push'
+      #     runs-on: ubuntu-latest
+      #     steps:
+      #       - uses: actions/checkout@v4
+      #       - run: npm run build
+prevention:
+  - "Only use paths: and paths-ignore: under push: and pull_request: trigger blocks"
+  - "Verify which filters apply to which events using the GitHub Actions syntax reference"
+  - "actionlint reports a warning when paths: is used under schedule: — run it in CI"
+  - "For scheduled conditional logic, gate using job if: conditions on in-workflow checks"
+docs:
+  - url: "https://docs.github.com/en/actions/writing-workflows/workflow-syntax-for-github-actions#onpushpull_requestpull_request_targetpaths"
+    label: "Workflow syntax: paths filter (applies to push and pull_request only)"
+  - url: "https://docs.github.com/en/actions/writing-workflows/choosing-when-your-workflow-runs/events-that-trigger-workflows#schedule"
+    label: "Events that trigger workflows: schedule"
+  - url: "https://github.com/dorny/paths-filter"
+    label: "dorny/paths-filter — detect file changes inside a scheduled workflow"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@htekdev/actions-debugger",
-  "version": "1.0.83",
+  "version": "1.0.84",
   "description": "65+ real GitHub Actions errors, queryable by agents. CLI + MCP server + Copilot skills + error database.",
   "type": "module",
   "main": "./dist/index.js",