Skip to content

[aw] Agentic Workflow Auto-Update failed #19626

Closed

Description

@github-actions

Workflow Failure

Workflow: Agentic Workflow Auto-Update
Branch: main
Run: https://github.com/dotnet/fsharp/actions/runs/24755189596
Pull Request: #15107

🛡️ Protected Files: The code push was refused because the patch modifies protected files (package manifests, agent instruction files, or repository security configuration). This protection guards against unintended supply chain changes.

Target Pull Request: #15107

Blocked Operations:

  • create_pull_request: Cannot create pull request: patch modifies protected files (.github/agents/agentic-workflows.agent.md, .github/aw/actions-lock.json, .github/workflows/aw-auto-update.lock.yml, .github/workflows/regression-pr-shepherd.lock.yml, .github/workflows/repo-assist.lock.yml). Add them to the allowed-files configuration field or set protected-files: fallback-to-issue to create a review issue instead.
⚙️ Configure protected-files: fallback-to-issue
safe-outputs:
  create-pull-request:
    protected-files: fallback-to-issue

Action Required

Assign this issue to Copilot using the agentic-workflows sub-agent to automatically debug and fix the workflow failure.

Debug with any coding agent

Use this prompt with any coding agent (GitHub Copilot, Claude, Gemini, etc.):

Debug the agentic workflow failure using https://raw.githubusercontent.com/github/gh-aw/main/debug.md

The failed workflow run is at https://github.com/dotnet/fsharp/actions/runs/24755189596
Manually invoke the agent

Debug this workflow failure using your favorite Agent CLI and the agentic-workflows prompt.

Tip

Stop reporting this workflow as a failure

To stop a workflow from creating failure issues, set report-failure-as-issue: false in its frontmatter:

safe-outputs:
  report-failure-as-issue: false

Generated from Agentic Workflow Auto-Update · ◷

  • expires on Apr 29, 2026, 1:29 AM UTC

Activity

  1. github-actions commented on Apr 23, 2026

    @github-actions
    ContributorAuthor

    Agent job 24811467562 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  2. github-actions commented on Apr 24, 2026

    @github-actions
    ContributorAuthor

    Agent job 24867358499 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  3. github-actions commented on Apr 25, 2026

    @github-actions
    ContributorAuthor

    Agent job 24919162885 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  4. github-actions commented on Apr 26, 2026

    @github-actions
    ContributorAuthor

    Agent job 24945152292 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  5. github-actions commented on Apr 27, 2026

    @github-actions
    ContributorAuthor

    Agent job 24972169444 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  6. github-actions commented on Apr 28, 2026

    @github-actions
    ContributorAuthor

    Agent job 25028721210 failed.

    🛡️ Protected Files: The code push was refused because the patch modifies protected files (package manifests, agent instruction files, or repository security configuration). This protection guards against unintended supply chain changes.

    Target Pull Request: #15107

    Blocked Operations:

    • create_pull_request: Cannot create pull request: patch modifies protected files (.github/agents/agentic-workflows.agent.md, .github/aw/actions-lock.json, .github/workflows/aw-auto-update.lock.yml, .github/workflows/regression-pr-shepherd.lock.yml, .github/workflows/repo-assist.lock.yml). Add them to the allowed-files configuration field or set protected-files: fallback-to-issue to create a review issue instead.
    ⚙️ Configure protected-files: fallback-to-issue
    safe-outputs:
      create-pull-request:
        protected-files: fallback-to-issue

    Generated from Agentic Workflow Auto-Update · ◷

  7. github-actions commented on Apr 29, 2026

    @github-actions
    ContributorAuthor

    Agent job 25086324428 failed.

    Generated from Agentic Workflow Auto-Update · ◷

  8. github-actions commented on Apr 30, 2026

    @github-actions
    ContributorAuthor

    Agent job 25142552834 failed.

    🔒 Lock File Out of Sync: The workflow could not start because its compiled lock file no longer matches the source markdown.

    This means the workflow's .md file was edited but gh aw compile was not run afterwards to regenerate the corresponding .lock.yml file. The agent is prevented from running against a stale configuration to avoid unexpected behaviour.

    To fix, recompile the workflow:

    gh aw compile

    Then commit and push the updated .lock.yml file.

    More ways to recompile

    Using the gh-aw MCP server (if configured):

    { "tool": "compile", "arguments": { "validate": true } }

    Recompile all workflows at once:

    gh aw compile --all

    Verify the result:

    gh aw compile --validate
    How to investigate the mismatch

    The workflow run logs contain a verbose debug pass that shows exactly what was hashed. Search the Check workflow lock file step logs for lines starting with [hash-debug] to see:

    • The raw frontmatter text that was used as input
    • Any imported files that were included in the hash
    • The canonical JSON that was fed to SHA-256
    • The resulting hash value

    This makes it easy to spot accidental whitespace changes, encoding differences, or import path drift.

    How to disable this check

    [!CAUTION]
    Disabling this check means the agent can run against an out-of-date compiled workflow. Only disable it if you have an alternative mechanism to keep lock files in sync.

    Set stale-check: false in the on: section of your workflow frontmatter:

    on:
      issues:
        types: [opened]
      stale-check: false

    After editing, recompile the workflow: gh aw compile

    Generated from Agentic Workflow Auto-Update · ◷

  9. github-actions commented on May 1, 2026

    @github-actions
    ContributorAuthor

    Agent job 25197848036 failed.

    🔒 Lock File Out of Sync: The workflow could not start because its compiled lock file no longer matches the source markdown.

    This means the workflow's .md file was edited but gh aw compile was not run afterwards to regenerate the corresponding .lock.yml file. The agent is prevented from running against a stale configuration to avoid unexpected behaviour.

    To fix, recompile the workflow:

    gh aw compile

    Then commit and push the updated .lock.yml file.

    More ways to recompile

    Using the gh-aw MCP server (if configured):

    { "tool": "compile", "arguments": { "validate": true } }

    Recompile all workflows at once:

    gh aw compile --all

    Verify the result:

    gh aw compile --validate
    How to investigate the mismatch

    The workflow run logs contain a verbose debug pass that shows exactly what was hashed. Search the Check workflow lock file step logs for lines starting with [hash-debug] to see:

    • The raw frontmatter text that was used as input
    • Any imported files that were included in the hash
    • The canonical JSON that was fed to SHA-256
    • The resulting hash value

    This makes it easy to spot accidental whitespace changes, encoding differences, or import path drift.

    How to disable this check

    [!CAUTION]
    Disabling this check means the agent can run against an out-of-date compiled workflow. Only disable it if you have an alternative mechanism to keep lock files in sync.

    Set stale-check: false in the on: section of your workflow frontmatter:

    on:
      issues:
        types: [opened]
      stale-check: false

    After editing, recompile the workflow: gh aw compile

    Generated from Agentic Workflow Auto-Update · ◷

  10. github-actions commented on May 2, 2026

    @github-actions
    ContributorAuthor

    Agent job 25240238537 failed.

    🔒 Lock File Out of Sync: The workflow could not start because its compiled lock file no longer matches the source markdown.

    This means the workflow's .md file was edited but gh aw compile was not run afterwards to regenerate the corresponding .lock.yml file. The agent is prevented from running against a stale configuration to avoid unexpected behaviour.

    To fix, recompile the workflow:

    gh aw compile

    Then commit and push the updated .lock.yml file.

    More ways to recompile

    Using the gh-aw MCP server (if configured):

    { "tool": "compile", "arguments": { "validate": true } }

    Recompile all workflows at once:

    gh aw compile --all

    Verify the result:

    gh aw compile --validate
    How to investigate the mismatch

    The workflow run logs contain a verbose debug pass that shows exactly what was hashed. Search the Check workflow lock file step logs for lines starting with [hash-debug] to see:

    • The raw frontmatter text that was used as input
    • Any imported files that were included in the hash
    • The canonical JSON that was fed to SHA-256
    • The resulting hash value

    This makes it easy to spot accidental whitespace changes, encoding differences, or import path drift.

    How to disable this check

    [!CAUTION]
    Disabling this check means the agent can run against an out-of-date compiled workflow. Only disable it if you have an alternative mechanism to keep lock files in sync.

    Set stale-check: false in the on: section of your workflow frontmatter:

    on:
      issues:
        types: [opened]
      stale-check: false

    After editing, recompile the workflow: gh aw compile

    Generated from Agentic Workflow Auto-Update · ◷

  11. github-actions commented on May 3, 2026

    @github-actions
    ContributorAuthor

    Agent job 25266592333 failed.

    🔒 Lock File Out of Sync: The workflow could not start because its compiled lock file no longer matches the source markdown.

    This means the workflow's .md file was edited but gh aw compile was not run afterwards to regenerate the corresponding .lock.yml file. The agent is prevented from running against a stale configuration to avoid unexpected behaviour.

    To fix, recompile the workflow:

    gh aw compile

    Then commit and push the updated .lock.yml file.

    More ways to recompile

    Using the gh-aw MCP server (if configured):

    { "tool": "compile", "arguments": { "validate": true } }

    Recompile all workflows at once:

    gh aw compile --all

    Verify the result:

    gh aw compile --validate
    How to investigate the mismatch

    The workflow run logs contain a verbose debug pass that shows exactly what was hashed. Search the Check workflow lock file step logs for lines starting with [hash-debug] to see:

    • The raw frontmatter text that was used as input
    • Any imported files that were included in the hash
    • The canonical JSON that was fed to SHA-256
    • The resulting hash value

    This makes it easy to spot accidental whitespace changes, encoding differences, or import path drift.

    How to disable this check

    [!CAUTION]
    Disabling this check means the agent can run against an out-of-date compiled workflow. Only disable it if you have an alternative mechanism to keep lock files in sync.

    Set stale-check: false in the on: section of your workflow frontmatter:

    on:
      issues:
        types: [opened]
      stale-check: false

    After editing, recompile the workflow: gh aw compile

    Generated from Agentic Workflow Auto-Update · ◷

  12. github-actions commented on May 4, 2026

    @github-actions
    ContributorAuthor

    Agent job 25296521601 failed.

    🔒 Lock File Out of Sync: The workflow could not start because its compiled lock file no longer matches the source markdown.

    This means the workflow's .md file was edited but gh aw compile was not run afterwards to regenerate the corresponding .lock.yml file. The agent is prevented from running against a stale configuration to avoid unexpected behaviour.

    To fix, recompile the workflow:

    gh aw compile

    Then commit and push the updated .lock.yml file.

    More ways to recompile

    Using the gh-aw MCP server (if configured):

    { "tool": "compile", "arguments": { "validate": true } }

    Recompile all workflows at once:

    gh aw compile --all

    Verify the result:

    gh aw compile --validate
    How to investigate the mismatch

    The workflow run logs contain a verbose debug pass that shows exactly what was hashed. Search the Check workflow lock file step logs for lines starting with [hash-debug] to see:

    • The raw frontmatter text that was used as input
    • Any imported files that were included in the hash
    • The canonical JSON that was fed to SHA-256
    • The resulting hash value

    This makes it easy to spot accidental whitespace changes, encoding differences, or import path drift.

    How to disable this check

    [!CAUTION]
    Disabling this check means the agent can run against an out-of-date compiled workflow. Only disable it if you have an alternative mechanism to keep lock files in sync.

    Set stale-check: false in the on: section of your workflow frontmatter:

    on:
      issues:
        types: [opened]
      stale-check: false

    After editing, recompile the workflow: gh aw compile

    Generated from Agentic Workflow Auto-Update · ◷

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Type

No type

Projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions