[daily secrets] Daily Secrets Analysis Report - 2026-09-17

[github-actions[bot]]

🔐 Daily Secrets Analysis Report

Date: 2026-09-17
Workflow Files Analyzed: 299
Run: §35254120691

📊 Executive Summary

• Total Secret References: 10,307 (secrets.*)
• GitHub Token References: 1,916 (github.token)
• Unique Secret Types: 41
• Redaction Coverage: 299/299 workflows (100%)
• Permission Blocks: 299/299 workflows (100%)
• Token Cascade Chains: 1,014 fallback instances (GH_AW_GITHUB_MCP_SERVER_TOKEN || GH_AW_GITHUB_TOKEN || GITHUB_TOKEN)

🛡️ Security Posture

✅ Redaction System: 299/299 workflows have redact_secrets steps — full coverage, no gaps.
✅ Token Cascades: 1,014 instances of the standard fallback chain, indicating consistent least-privilege token selection across workflows.
✅ Permission Blocks: 299/299 workflows declare explicit permissions: blocks — no workflow relies on implicit default permissions.
✅ Template Injection Check (TestCompiledLockFiles_NoGitHubEventExpressionsInRunScripts): PASSED — no direct github.event.* interpolation found in executable run: scripts.
✅ Secrets-in-Outputs Check (TestCompiledLockFiles_NoSecretsInOutputs): PASSED — no secret references found in job or workflow_call outputs.

🎯 Key Findings

1. Complete redaction and permission coverage: Every one of the 299 compiled workflows has both a redaction step and an explicit permissions block — no regressions detected in these two structural controls.
2. Token usage is heavily concentrated: GITHUB_TOKEN (5,348 refs) and GH_AW_GITHUB_TOKEN (4,557 refs) together account for ~96% of all secret references, reflecting the standard GitHub API/MCP access pattern used by nearly all workflows.
3. Observability secrets are a large secondary cluster: OTEL/Sentry/Grafana endpoint and authorization secrets (GH_AW_OTEL_SENTRY_AUTHORIZATION, GH_AW_OTEL_SENTRY_ENDPOINT, GH_AW_OTEL_GRAFANA_AUTHORIZATION, GH_AW_OTEL_GRAFANA_ENDPOINT) total ~1,984 references, showing broad telemetry instrumentation across workflows.
4. AI provider keys are well-distributed: OPENAI_API_KEY (398), CODEX_API_KEY (397), and ANTHROPIC_API_KEY (255) show multi-engine support is actively exercised rather than concentrated on a single provider.
5. No template injection or output-leak anomalies: Both automated compiler-level security tests passed cleanly against the full set of 299 lock files.

💡 Recommendations

1. Maintain the 100% redaction/permissions baseline — treat any future compile that drops a workflow below full coverage on either metric as a release-blocking regression.
2. Continue running the two Go security tests (NoGitHubEventExpressionsInRunScripts, NoSecretsInOutputs) in CI on every PR touching .lock.yml generation logic, since they are the deterministic guardrails against the two highest-impact secret risks (template injection, output leakage).
3. Track the token cascade ratio over time — 1,014 cascade instances against 299 workflows (~3.4 per workflow) is a healthy baseline; a sudden drop could indicate workflows bypassing the fallback pattern in favor of a single hardcoded token.

🔑 Top 10 Secrets by Usage

Rank	Secret Name	Occurrences	Type
1	GITHUB_TOKEN	5,348	GitHub Token
2	GH_AW_GITHUB_TOKEN	4,557	GitHub Token
3	GH_AW_GITHUB_MCP_SERVER_TOKEN	1,945	GitHub Token
4	GH_AW_OTEL_SENTRY_AUTHORIZATION	744	Observability
5	COPILOT_GITHUB_TOKEN	527	GitHub Token
6	GH_AW_OTEL_SENTRY_ENDPOINT	497	Observability
7	GH_AW_OTEL_GRAFANA_AUTHORIZATION	495	Observability
8	OPENAI_API_KEY	398	AI Provider
9	CODEX_API_KEY	397	AI Provider
10	ANTHROPIC_API_KEY	255	AI Provider

📈 Trends

No historical baseline is currently available for this repository — this is the first recorded run of the daily secrets analysis. Today's statistics establish the baseline for future day-over-day comparisons:

• Total secret references: 10,307
• Unique secret types: 41
• Redaction coverage: 100%
• Permission block coverage: 100%

📖 Reference Documentation

For detailed information about secret usage patterns, see:

• Redaction System: actions/setup/js/redact_secrets.cjs
• Compiler security tests: pkg/workflow/ (TestCompiledLockFiles_NoGitHubEventExpressionsInRunScripts, TestCompiledLockFiles_NoSecretsInOutputs)

---

Generated: 2026-09-17 17:44 UTC
Workflow: Daily Secrets Analysis Agent

Warning

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

• api.github.com

[!TIP]
api.github.com is blocked because GitHub API access uses the built-in GitHub tools by default. Instead of adding api.github.com to network.allowed, use tools.github.mode: gh-proxy for direct pre-authenticated GitHub CLI access without requiring network access to api.github.com:

tools:
  github:
    mode: gh-proxy

See GitHub Tools for more information on gh-proxy mode.

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "api.github.com"

See Network Configuration for more information.

Generated by 🔒 Daily Secrets Analysis Agent · copilot · auto · 52.6 AIC · ⌖ 6.84 AIC · ⊞ 9.7K · ◷

• expires on Sep 20, 2026, 9:48 AM UTC-08:00