Workflow Skill Extractor Report — 2026-03-26

[github-actions[bot]]

This report identifies reusable skills and shared component opportunities across the agentic workflow repository, based on analysis of 178 workflows and 64 existing shared components.

Key Statistics

Metric	Value
Total workflows analyzed	178
Existing shared components	64
New skills identified	6
High-priority recommendations	3
Estimated total lines saved	~2,800+

The highest-impact finding: 80% of all workflows (143/178) manually embed the exact same 8-line noop reminder that could be extracted into a single shared component.

---

Workflows Analyzed (Sample)

All 178 .github/workflows/*.md files were analyzed. Representative sample includes:

daily-issues-report.md, weekly-issue-summary.md, auto-triage-issues.md, issue-triage-agent.md, daily-performance-summary.md, audit-workflows.md, daily-firewall-report.md, agent-performance-analyzer.md, agentic-observability-kit.md, metrics-collector.md, safe-output-health.md, daily-safe-output-optimizer.md, daily-code-metrics.md, copilot-agent-analysis.md, security-review.md, daily-security-red-team.md, daily-malicious-code-scan.md, repository-quality-improver.md, workflow-health-manager.md

Existing Shared Components (Top Imports by Frequency)

Component	Workflows Using It
shared/reporting.md	~60+
shared/trends.md	~25+
shared/python-dataviz.md	~20+
shared/jqschema.md	~15+
shared/issues-data-fetch.md	~10+
shared/copilot-pr-analysis-base.md	~5+
shared/aw-logs-24h-fetch.md	2

Engine distribution: copilot (83), codex (36), claude (25), gemini (4), none/custom (30)

---

High Priority Skills

1. Noop Reminder Prompt

• Frequency: 143 workflows (80% of all workflows)
• Size: ~8 lines per workflow → ~1,144 lines total
• Priority: 🔴 High
• Description: An exact 8-line block telling agents to call noop when no action is needed. The text is identical in 138 workflows and near-identical in 5 more.
• Workflows: Nearly every workflow in the repository
• Recommendation: Extract to shared/noop-reminder.md
• Issue: [refactoring] Extract noop reminder into shared component #23099

2. Agentic-Workflows MCP Tool Configuration

• Frequency: 28 workflows
• Size: ~15–25 lines guidance per workflow → ~500 lines total
• Priority: 🔴 High
• Description: The agentic-workflows: MCP tool is declared identically across 28 workflows, each also independently documenting how to use the status, logs, and audit MCP tools. Only 2 of these import the narrow shared/aw-logs-24h-fetch.md; 26 have no shared foundation.
• Workflows: audit-workflows.md, daily-firewall-report.md, agent-performance-analyzer.md, agentic-observability-kit.md, metrics-collector.md, daily-observability-report.md, portfolio-analyst.md, prompt-clustering-analysis.md, cloclo.md, craft.md, q.md, deep-report.md, dev-hawk.md, mcp-inspector.md and 14 more
• Recommendation: Extract to shared/agentic-workflows-tool.md
• Issue: [refactoring] Extract agentic-workflows MCP tool configuration into shared component #23102

3. Repo-Memory Configuration Guidelines

• Frequency: 26 workflows
• Size: ~20–40 lines guidance per workflow → ~600 lines total
• Priority: 🟠 Medium-High
• Description: 26 workflows use repo-memory: with near-identical configuration (same file-glob patterns, same max-file-size: 102400) and duplicate prompt instructions for JSON Lines format, 90-day data retention, and file structure.
• Workflows: agent-performance-analyzer.md, audit-workflows.md, copilot-agent-analysis.md, copilot-pr-nlp-analysis.md, copilot-session-insights.md, daily-code-metrics.md, daily-copilot-token-report.md, metrics-collector.md, workflow-health-manager.md, deep-report.md and 16 more
• Recommendation: Extract to shared/repo-memory-guidelines.md
• Issue: [refactoring] Extract repo-memory configuration patterns into shared component #23101

---

Medium Priority Skills

4. Daily Audit Discussion Pattern

• Frequency: ~40 workflows
• Description: Many daily workflows have nearly identical safe-outputs blocks (create-discussion: expires: 1d, category: "audits", max: 1, close-older-discussions: true, upload-asset:). While frontmatter merging makes full extraction complex, a documentation guide would help.
• Priority: 🟡 Medium — frontmatter values vary (title-prefix, expires duration), making clean extraction harder

5. Trending Charts + Agentic-Workflows Combo

• Frequency: 4–5 workflows
• Description: audit-workflows.md, daily-firewall-report.md, portfolio-analyst.md, prompt-clustering-analysis.md all combine agentic-workflows: + shared/trending-charts-simple.md + shared/reporting.md for the "download workflow logs → analyze → generate charts" pattern
• Priority: 🟡 Medium — already partially addressed by individual imports; a combined shared/workflow-log-analysis.md would save 3–5 import lines each

6. Parallel Agentic Workflow Status Check

• Frequency: ~5 workflows
• Description: Multiple meta-orchestrator workflows (agent-performance-analyzer.md, workflow-health-manager.md, metrics-collector.md) check workflow status using identical MCP status call patterns and store results in identical repo-memory structures under memory/meta-orchestrators/
• Priority: 🟡 Medium — the overlap is specific to the meta-orchestrator cluster

---

Detailed Recommendations

Recommendation 1: shared/noop-reminder.md

Full Details

Current State (identical in 138 workflows):

**Important**: If no action is needed after completing your analysis, you **MUST** call the `noop` safe-output tool with a brief explanation. Failing to call any safe-output tool is the most common cause of safe-output workflow failures.

\`\`\`json
{"noop": {"message": "No action needed: [brief explanation of what was analyzed and why]"}}
\`\`\`

Proposed Shared Component (.github/workflows/shared/noop-reminder.md):

---
---

**Important**: If no action is needed after completing your analysis, you **MUST** call the `noop` safe-output tool with a brief explanation. Failing to call any safe-output tool is the most common cause of safe-output workflow failures.

\`\`\`json
{"noop": {"message": "No action needed: [brief explanation of what was analyzed and why]"}}
\`\`\`

Migration Path:

1. Create shared/noop-reminder.md
2. Search-and-replace the 8-line block across all 143 affected workflows
3. Add - shared/noop-reminder.md to their imports: section (or create one)
4. Normalize the 5 workflows with variant wording
5. make recompile

Impact:

• Lines saved: ~1,144 lines (138 × 8 + 5 × ~8)
• Single update point for wording changes
• Eliminates 5 inconsistent variants

Recommendation 2: shared/agentic-workflows-tool.md

Full Details

Current State (example from audit-workflows.md):

tools:
  agentic-workflows:
  ...
```

Plus inline prompt text like:
```
Use gh-aw MCP server (not CLI directly). Run `status` tool to verify.
**Collect Logs**: Use MCP `logs` tool to download workflow logs:
Use the agentic-workflows MCP tool `logs` with parameters:
- start_date: "-1d" (last 24 hours)

This guidance is duplicated with slight variations across 28 workflows.

Proposed Shared Component (.github/workflows/shared/agentic-workflows-tool.md):

---
tools:
  agentic-workflows:
  cache-memory: true
---

## Agentic Workflows MCP Tool

The `agentic-workflows` MCP server is pre-configured. Use its tools — **not** the `./gh-aw` CLI binary — to query workflow data:

- **`status`**: List all workflows and recent run status
- **`logs`**: Download workflow run logs (`start_date`, `engine`, `output` parameters)
- **`audit`**: Audit a specific run (`run_id` parameter)

Logs are stored per-run in `output/run-(id)/` with `aw_info.json`, `activation/`, and `agent/` subdirectories.

Migration Path:

1. Create shared/agentic-workflows-tool.md
2. Add - shared/agentic-workflows-tool.md to the imports: of all 28 affected workflows
3. Remove the agentic-workflows: from each workflow's tools: section (now provided by the import)
4. Remove duplicate MCP usage guidance from each workflow's body
5. make recompile and validate 3–5 workflows

Impact:

• Lines saved: ~500 lines across 28 workflows
• When the MCP tool interface evolves, 1 update propagates to all 28 workflows
• Standardizes agentic-workflows: vs agentic-workflows: true variance

Recommendation 3: shared/repo-memory-guidelines.md

Full Details

Current State (repeated across 26 workflows, with variations):

repo-memory:
  branch-name: memory/copilot-agent-analysis
  description: "Historical agent performance metrics"
  file-glob: ["memory/copilot-agent-analysis/*.json", "memory/copilot-agent-analysis/*.jsonl",
              "memory/copilot-agent-analysis/*.csv", "memory/copilot-agent-analysis/*.md"]
  max-file-size: 102400  # 100KB

Note: max-file-size: 102400 is identical in all 26 workflows. file-glob follows the same ["*.json", "*.jsonl", "*.csv", "*.md"] pattern in most.

Proposed Shared Component (.github/workflows/shared/repo-memory-guidelines.md):

---
---

## Repo-Memory Data Storage Guidelines

Configure `repo-memory` in your workflow's `tools:` with a unique branch and glob:

\`\`\`yaml
tools:
  repo-memory:
    branch-name: memory/<your-workflow-name>
    description: "Historical <metric type> data"
    file-glob: ["memory/<your-workflow-name>/*.json", "memory/<your-workflow-name>/*.jsonl",
                "memory/<your-workflow-name>/*.csv", "memory/<your-workflow-name>/*.md"]
    max-file-size: 102400  # 100KB
\`\`\`

### Storage Format: JSON Lines

Append time-series data points using ISO 8601 timestamps:

\`\`\`python
import json
from datetime import datetime
with open('/tmp/gh-aw/repo-memory/default/memory/<workflow>/history.jsonl', 'a') as f:
    f.write(json.dumps({"timestamp": datetime.utcnow().isoformat() + "Z", "value": ...}) + '\n')
\`\`\`

### 90-Day Retention (Required)

\`\`\`python
from datetime import datetime, timedelta
import pandas as pd
cutoff = datetime.utcnow() - timedelta(days=90)
df = pd.read_json('history.jsonl', lines=True)
df = df[pd.to_datetime(df['timestamp']) >= cutoff]
\`\`\`

Migration Path:

1. Create shared/repo-memory-guidelines.md
2. Add - shared/repo-memory-guidelines.md to the 26 affected workflows
3. Remove duplicated JSON Lines / retention guidance from workflow bodies
4. Standardize all workflows to 90-day retention
5. make recompile

Impact:

• Lines saved: ~600 lines across 26 workflows
• Eliminates retention-policy inconsistency (some use 30-day, some 90-day, some none)
• Single update point for storage best practices

---

Impact Analysis

By Category

Category	Skills	Lines Saved	Workflows Affected
Prompt Skills	2 (noop, repo-memory guidelines)	~1,744	143 + 26
Tool Configurations	1 (agentic-workflows)	~500	28
Combo Patterns	2 (trending+aw, meta-orchestrator)	~200	8
Total	5	~2,444	varies

By Priority

Priority	Skills	Lines Saved	Workflows Affected
High	2	~1,644	143, 28
Medium-High	1	~600	26
Medium	3	~200	4–8

---

Created Issues

This analysis created the following actionable issues:

1. [refactoring] Extract noop reminder into shared component #23099 — Extract noop reminder into shared component
2. [refactoring] Extract agentic-workflows MCP tool configuration into shared component #23102 — Extract agentic-workflows MCP tool configuration into shared component
3. [refactoring] Extract repo-memory configuration patterns into shared component #23101 — Extract repo-memory configuration patterns into shared component

---

Next Steps

1. Review and prioritize the 3 created issues
2. Start with noop-reminder (highest impact, easiest migration — text search/replace)
3. Continue with agentic-workflows-tool (high impact, moderate migration)
4. Tackle repo-memory-guidelines after the above are stable
5. Consider automating detection: add a CI check that warns when the noop reminder text is manually embedded instead of imported
6. Schedule a follow-up extractor run in ~1 month after migrations are complete

---

Methodology

• Analyzed 178 workflow markdown files in .github/workflows/
• Reviewed 64 existing shared components in .github/workflows/shared/
• Applied pattern recognition across 5 categories: tool configs, prompt skills, data processing, setup steps, shared data operations
• Quantified impact by counting exact and near-exact string matches
• Prioritized based on frequency, line savings, and implementation complexity

Workflow Run: §23591960334
Analysis Date: 2026-03-26
Analyzer: Workflow Skill Extractor v1.0

AI generated by Workflow Skill Extractor · history

• expires on Apr 2, 2026, 11:43 AM UTC