Problem Description
The issue-monster workflow has been failing on every single run (5+ consecutive failures in the last 24 hours). The workflow runs on a schedule every 30 minutes, generating continuous noise and never accomplishing its mission.
Tool
- Tool:
audit, logs
- Workflow: Issue Monster (
.github/workflows/issue-monster.md)
Failure Details
All runs fail in the agent job at the "Validate lockdown mode requirements" step.
Error message:
##[error]Lockdown mode is enabled (lockdown: true) but no custom GitHub token is configured.
Please configure one of the following as a repository secret:
- GH_AW_GITHUB_TOKEN (recommended)
- GH_AW_GITHUB_MCP_SERVER_TOKEN (alternative)
- Custom github-token in your workflow frontmatter
See: https://github.com/github/gh-aw/blob/main/docs/src/content/docs/reference/auth.mdx
Root Cause
The issue-monster.md workflow has:
tools:
github:
lockdown: true
toolsets: [default, pull_requests]
The lockdown: true setting requires a custom GitHub token (GH_AW_GITHUB_TOKEN or GH_AW_GITHUB_MCP_SERVER_TOKEN) to be set as a repository secret. This secret is not configured.
Affected Runs (last 24 hours)
Steps to Reproduce
- Look at recent Issue Monster workflow runs via
logs tool: workflow_name: "issue-monster"
- All runs show
conclusion: failure and error_count: 1
- Audit any run: the agent job fails at "Validate lockdown mode requirements" step
Expected Behavior
Either:
- Configure
GH_AW_GITHUB_TOKEN repository secret to allow lockdown mode to work, OR
- Remove
lockdown: true from the workflow if a custom token isn't available
Actual Behavior
The workflow fails immediately in the agent job, wastes ~2 minutes per run, and runs every 30 minutes.
Environment
- Repository: github/gh-aw
- Run ID: 22375263195 (Daily CLI Tools Exploratory Tester)
- Date: 2026-02-24
- gh-aw version: v0.0.415 (agent_version from aw_info.json)
Impact
- Severity: High
- Frequency: Always (100% failure rate, every 30 minutes)
- Workaround: Disable the workflow or remove
lockdown: true
Secondary Findings (from same testing session)
-
audit with invalid run ID returns a generic MCP error (-32603: failed to audit workflow run: exit status 1) instead of a user-friendly message like "Run not found". Consider improving the error message.
-
compile with actionlint: true, zizmor: true, or poutine: true fails with "docker images not ready" when Docker daemon is not running. The error is correct but could be more descriptive (e.g., "Docker daemon is not available - start Docker to use scan tools").
Generated by Daily CLI Tools Exploratory Tester
Problem Description
The
issue-monsterworkflow has been failing on every single run (5+ consecutive failures in the last 24 hours). The workflow runs on a schedule every 30 minutes, generating continuous noise and never accomplishing its mission.Tool
audit,logs.github/workflows/issue-monster.md)Failure Details
All runs fail in the
agentjob at the "Validate lockdown mode requirements" step.Error message:
Root Cause
The
issue-monster.mdworkflow has:The
lockdown: truesetting requires a custom GitHub token (GH_AW_GITHUB_TOKENorGH_AW_GITHUB_MCP_SERVER_TOKEN) to be set as a repository secret. This secret is not configured.Affected Runs (last 24 hours)
Steps to Reproduce
logstool:workflow_name: "issue-monster"conclusion: failureanderror_count: 1Expected Behavior
Either:
GH_AW_GITHUB_TOKENrepository secret to allow lockdown mode to work, ORlockdown: truefrom the workflow if a custom token isn't availableActual Behavior
The workflow fails immediately in the agent job, wastes ~2 minutes per run, and runs every 30 minutes.
Environment
Impact
lockdown: trueSecondary Findings (from same testing session)
auditwith invalid run ID returns a generic MCP error (-32603: failed to audit workflow run: exit status 1) instead of a user-friendly message like "Run not found". Consider improving the error message.compilewithactionlint: true,zizmor: true, orpoutine: truefails with"docker images not ready"when Docker daemon is not running. The error is correct but could be more descriptive (e.g., "Docker daemon is not available - start Docker to use scan tools").