Workflow Skill Extraction Report: 3 shared-component adoption opportunities

[github-actions[bot]]

Workflow Skill Extraction Report

Methodology

Analyzed /tmp/gh-aw/agent/workflow-index.json (266 workflows total, 77 existing shared components in .github/workflows/shared/) to map import frequency, engines, and tool usage. Cross-referenced high-frequency patterns against existing shared components to distinguish "needs a new component" from "component exists but is under-adopted." Deep-dived candidate workflows with grep/glob to confirm concrete duplication evidence (line numbers, package-install patterns, pricing config) before finalizing recommendations.

Coverage summary

• 266 total workflows, 34 of which live in shared/
• Import frequency leaders: shared/otlp.md (228), shared/reporting.md (53), shared/github-guard-policy.md (19), shared/gh.md (12)
• 116 workflows have engine: null (likely default/copilot), 98 explicitly copilot, 21 pi, 21 claude, 7 codex, 1 each antigravity/gemini/opencode
• Most core infra (OTLP, reporting) is already well-shared; the biggest opportunities are adoption gaps for components that already exist but aren't used everywhere.

Opportunities identified (priority order)

#	Priority	Skill	Workflows affected	Issue
1	High	Adopt shared/python-dataviz.md for matplotlib/pandas/seaborn setup	12	#aw_pydv1
2	Medium	Adopt shared/default-ai-credits-pricing.md to prevent pricing failures	~30 (non-copilot engines), broader potential	#aw_price2
3	Low	Consolidate manual MCP pagination guidance via shared/mcp-pagination.md	~21	#aw_pgn3

Details

1. Python data-viz setup (High priority)

12 workflows (agentic-token-audit.md, agentic-token-trend-audit.md, api-consumption-report.md, daily-code-metrics.md, daily-experiment-report.md, daily-firewall-report.md, daily-issues-report.md, daily-news.md, daily-performance-summary.md, daily-repo-chronicle.md, prompt-clustering-analysis.md, weekly-issue-summary.md) use matplotlib/pandas/seaborn without importing the existing shared/python-dataviz.md or shared/charts-with-trending.md. Setup approaches diverge inconsistently (pip install --target, python3 -m venv, pypy3 -m pip install), risking version drift and re-introducing solved problems (e.g. --break-system-packages). Estimated 150-300 duplicated lines removable, plus consistency of a single venv/package strategy.

2. Default AI-credits pricing fallback (Medium priority)

Only 5 of 266 workflows import shared/default-ai-credits-pricing.md. This is a latent-risk gap rather than visible duplication: workflows using less-common engines (pi: 21, codex: 7, antigravity/gemini/opencode: 1 each) are most exposed to missing_model_pricing/unknown_model_ai_credits failures when a model lacks catalog pricing. Adding the 1-line import is cheap insurance against a class of intermittent, hard-to-debug failures.

3. MCP pagination guidance consolidation (Low priority)

21 workflows reference per_page/pagination manually without importing shared/mcp-pagination.md, leading to inconsistent perPage thresholds and repeated prose about the 25,000-token MCP response limit. Lower priority since duplication is prose (not executable code) and the highest-risk bulk-list cases are already covered by shared/github-mcp-pagination-wrappers.md.

Impact summary

• High-impact, low-risk: python-dataviz adoption reduces real duplicated setup code across 12 workflows with a clear, already-built target.
• Reliability hardening: default-ai-credits-pricing adoption is cheap and directly prevents recurring pipeline failures on non-mainstream engines.
• Consistency polish: MCP pagination guidance consolidation is lower urgency but easy to bundle into any of the above migrations.

No new shared components are proposed — all three recommendations are adoption/migration tasks for components that already exist in .github/workflows/shared/, which keeps implementation risk low (no new abstractions to validate) while still meaningfully reducing duplication and failure surface.

References:

• Created issues: see linked issues above (python-dataviz adoption, default-ai-credits-pricing adoption, MCP pagination consolidation)

Generated by 🔍 Workflow Skill Extractor · aut00 · 63.5 AIC · ⌖ 5.58 AIC · ⊞ 6.8K · ◷

• expires on Aug 5, 2026, 8:47 PM UTC-08:00

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-08-06T04:47:00.868Z.

Closed by Workflow

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-08-06T04:47:00.868Z.

Closed by Workflow

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-08-06T04:47:00.868Z.

Closed by Workflow