Agentic Workflow Audit — 2026-03-11 (Run #233)

[github-actions[bot]]

Daily audit of agentic workflows as of 2026-03-11. This is the first audit run (#233) establishing baseline metrics. Note: the logs MCP tool was unavailable (gh CLI not authenticated), so detailed per-run metrics (token usage, cost, errors) could not be collected. Analysis is based on workflow status only.

---

Summary

Metric	Value
Total workflows registered	166
Active workflows	35
Unknown status (event-triggered)	131
Expired workflows	2
AWF version	v0.23.0
Firewall	Enabled (squid)

---

Engine Distribution

Engine	Workflows	%
Copilot	114	68.7%
Claude	40	24.1%
Codex	11	6.6%
Gemini	1	0.6%

---

Active Workflows (35)

View all active workflows

Workflow	Engine
artifacts-summary	copilot
audit-workflows	claude
ci-doctor ⚠️ Expired	copilot
cli-version-checker	claude
commit-changes-analyzer	claude
copilot-agent-analysis	claude
daily-doc-updater	claude
daily-news	copilot
dev	copilot
dev-hawk	copilot
duplicate-code-detector	codex
example-workflow-analyzer	claude
github-mcp-tools-report	claude
go-logger	claude
go-pattern-detector	claude
lockfile-stats	claude
mcp-inspector	copilot
mergefest	copilot
notion-issue-summary	copilot
pdf-summary	copilot
plan	copilot
poem-bot	copilot
q	copilot
repo-tree-map	copilot
research	copilot
scout	claude
semantic-function-refactor	claude
smoke-claude	claude
smoke-codex	codex
smoke-copilot	copilot
technical-doc-writer	copilot
terminal-stylist	copilot
unbloat-docs	claude
video-analyzer	copilot
weekly-issue-summary	copilot

---

Issues Found

⚠️ Expired Workflows (2)

Two workflows have exceeded their stop-after deadline and are marked Expired:

• ci-doctor — triggers on workflow_run (CI completes), stop-after: +1mo
• daily-team-status — daily schedule, stop-after: +1mo

These workflows should be reviewed and either renewed (update stop-after) or disabled if no longer needed.

⚠️ Logs Tool Unavailable

The logs MCP tool failed with exit status 1 (gh CLI not authenticated). This prevented:

• Per-run error analysis
• Token usage and cost tracking
• MCP failure detection
• Performance trend analysis

This is a recurring infrastructure constraint for this audit workflow. Future runs will benefit if gh CLI authentication is available.

ℹ️ Repo Memory Bootstrap Issue

This is the first audit run — no historical baseline exists in repo memory. The push_repo_memory validation reported an oversized diff (~26–27 KB) due to git pack format overhead when initializing a new branch for the first time. Historical comparison will become available in subsequent runs.

---

Observations

• 131 workflows show Unknown status — this is expected for event-triggered workflows (slash commands, PR labels, reactions) that haven't run recently or are disabled.
• The repository has a rich and diverse set of 166 agentic workflows across 4 AI engines (Copilot, Claude, Codex, Gemini).
• Claude-based workflows make up 24% of total and 17/35 (49%) of active workflows.
• All compiled workflows are up to date (compiled: Yes for all 166).

---

Recommendations

1. Renew or retire ci-doctor and daily-team-status — both have expired stop-after limits
2. Investigate logs tool — determine if gh CLI auth can be provided to enable full telemetry
3. Establish baseline — next run will have historical comparison data in repo memory

---

References:

• §22938628434 — this audit run

AI generated by Agentic Workflow Audit Agent · history

• expires on Mar 12, 2026, 5:49 AM UTC

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-03-12T05:49:52.304Z.

Closed by Workflow