Artifacts Usage Report

[github-actions[bot]]

Analysis window: recent runs (last ~1 hour of activity captured in this run — GitHub's Artifacts API only exposes non-expired artifacts, and the most recent runs dominate the sample). 20 workflows had artifacts across 20 distinct runs, totaling 99 artifacts, ~~135 MB combined.

Workflow Name	Artifacts Count	Total Size	Avg Size	Latest Run	Status
Daily Cli Tools Tester	8	9.3 MB	1.2 MB	2026-08-02 06:18 UTC	Active
Contribution Check	7	9.2 MB	1.3 MB	2026-08-02 05:54 UTC	Active
Schema Consistency Checker	7	9.2 MB	1.3 MB	2026-08-02 06:23 UTC	Active
Multi-Device Docs Tester	2	9.2 MB	4.6 MB	2026-08-02 06:24 UTC	Active
CLI Version Checker	8	9.1 MB	1.1 MB	2026-08-02 06:25 UTC	Active
Static Analysis Report	8	9.0 MB	1.1 MB	2026-08-02 06:19 UTC	Active
Daily Sub-Agent Model Resolution Audit	3	8.9 MB	3.0 MB	2026-08-02 06:25 UTC	Active
Issue Arborist	7	8.9 MB	1.3 MB	2026-08-02 06:19 UTC	Active
Daily Windows Terminal Integration Builder	7	8.8 MB	1.3 MB	2026-08-02 05:49 UTC	Active
Dictation Prompt Generator	5	8.7 MB	1.7 MB	2026-08-02 06:24 UTC	Active
GitHub Remote MCP Authentication Test	6	8.6 MB	1.4 MB	2026-08-02 06:23 UTC	Active
The Great Escapi	8	8.6 MB	1.1 MB	2026-08-02 05:48 UTC	Active
PR Sous Chef	5	8.5 MB	1.7 MB	2026-08-02 06:24 UTC	Active
AI Moderator	6	8.4 MB	1.4 MB	2026-08-02 05:57 UTC	Active
ESLint Refiner	1	8.3 MB	8.3 MB	2026-08-02 06:10 UTC	Active
Artifacts Usage Report	1	8.3 MB	8.3 MB	2026-08-02 06:23 UTC	Active
Daily Container Image Security Scan	1	8.3 MB	8.3 MB	2026-08-02 06:20 UTC	Active
Workflow Health Manager - Meta-Orchestrator	6	1.4 MB	243 KB	2026-08-02 05:43 UTC	Active
Safe Output Health Monitor	3	8.3 KB	2.8 KB	2026-08-02 05:41 UTC	Active
Agentic Maintenance	1	1.1 KB	1.1 KB	2026-08-02 05:41 UTC	Active

Methodology & data limitations

• Data source: GET /repos/github/gh-aw/actions/artifacts (single page, most recent 99 non-expired artifacts returned by the API at query time — GitHub does not retain historical/expired artifact metadata beyond expiry, so this reflects a live snapshot rather than a true rolling 30-day history).
• Each artifact was mapped to its originating run via run.workflow_run.id, then each run was resolved to its workflow name via GET /repos/github/gh-aw/actions/runs/{id}.
• All 20 workflows observed in this snapshot have zero expired artifacts, so all are marked Active. A true 30-day trend requires either enabling artifact-retention audit logging or running this report periodically and diffing snapshots, since expired artifacts disappear from the API.
• Sizes shown are size_in_bytes converted to MB/KB (1024-based).

Insights & Recommendations

• Most artifacts generated: Daily Cli Tools Tester, CLI Version Checker, and Static Analysis Report (8 artifacts each) — these run frequently (daily/hourly cadence) and produce multiple small artifacts (logs, agent output, usage stats) per run.
• Largest storage consumers: Several single-artifact workflows (ESLint Refiner, Artifacts Usage Report itself, Daily Container Image Security Scan) each upload one ~8.3 MB artifact, likely full container-build or docker-build cache artifacts rather than logs.
• Most efficient: Safe Output Health Monitor and Agentic Maintenance produce tiny (KB-scale) artifacts — good practice, minimal storage footprint.
• Trend: This snapshot only captures a short recent window because GitHub's Artifacts API surfaces only non-expired artifacts; most gh-aw lock workflows use short retention windows, so long-term trend analysis isn't possible from a single snapshot.

Recommendations:

1. For workflows uploading large (~8 MB) single artifacts (docker build caches, container scan results), consider shortening retention-days further or excluding these from artifact upload if not needed for debugging.
2. For high-count/low-size artifact workflows (Daily Cli Tools Tester, CLI Version Checker, Static Analysis Report), consider consolidating multiple small artifacts (e.g., agent, usage, aic-usage-cache) into a single combined archive per run to reduce artifact-count overhead.
3. To get true 30-day trends, schedule this report to run periodically and persist historical stats (e.g., in a tracking issue/discussion) rather than relying on a single point-in-time query, since expired artifacts are not retrievable retroactively.

Generated by 📦 Artifacts Usage Report · auto · 29.3 AIC · ⌖ 12 AIC · ⊞ 6K · ◷

• expires on Aug 2, 2026, 10:30 PM UTC-08:00