🔍 Static Analysis Report - February 11, 2026

[github-actions[bot]]

Analysis Summary

Daily static analysis scan of all agentic workflows using three security and code quality tools: zizmor (security), poutine (supply chain), and actionlint (linting with shellcheck).

• Tools Used: zizmor, poutine, actionlint (with shellcheck)
• Total Findings: 334
• Workflows Scanned: 149
• Workflows with Security Issues: 2 (1 medium, 1 low)
• Overall Security Posture: ✅ Good (no critical or high severity issues)

Findings by Tool

Tool	Total	Critical	High	Medium	Low	Info/Style
zizmor (security)	2	0	0	1	1	0
poutine (supply chain)	10	0	0	0	0	10
actionlint (shellcheck)	322	0	0	0	0	322
TOTAL	334	0	0	1	1	332

Key Insight: 96.4% of findings are informational/style issues from shellcheck. Only 2 security findings (both localized to specific workflows).

---

Critical Security Findings

🟡 Medium Severity (1 finding)

artipacked - Credential Persistence Through Artifacts

• Workflow: daily-copilot-token-report.lock.yml
• Location: Line 115
• Description: Credentials may persist through GitHub Actions artifacts
• Impact: Potential credential exposure if artifacts are publicly accessible
• Reference: (docs.zizmor.sh/redacted)
• Status: Recurring since Feb 5

🔵 Low Severity (1 finding)

template-injection - Code Injection via Template Expansion

• Workflow: mcp-inspector.lock.yml
• Location: Line 538
• Description: Potential code injection through template expansion
• Impact: Low risk if inputs are controlled
• Reference: (docs.zizmor.sh/redacted)
• Status: Recurring since Feb 5

---

Supply Chain Security Findings (Poutine)

⚠️ Warning Level (1 finding)

Issue Type	Count	Affected Workflow
default_permissions_on_risky_events	1	ai-moderator.lock.yml

Note: This is a significant improvement from 14 workflows on Feb 9 to just 1 workflow currently. ✅

ℹ️ Informational (9 findings)

Issue Type	Count	Description
unverified_script_exec	4	curl
github_action_from_unverified_creator_used	3	Actions from unverified creators
unpinnable_action	2	CI components that cannot be pinned

Assessment: These are informational notices about setup patterns. The curl | sh pattern is used for installing uv and gh-aw CLI tools, which is intentional and commonly used for CLI installations.

---

Code Quality Findings (Actionlint/Shellcheck)

📋 Style Issues (322 findings)

Issue Code	Count	Severity	Description
SC2129	164	style	Consider using grouped redirects: { cmd1; cmd2; } >> file
SC1003	158	info	Single quote escaping suggestion in domain strings

Impact: These are style recommendations in compiler-generated code, not security vulnerabilities.

Affected: All 149 compiled workflows contain these patterns.

---

Detailed Findings by Workflow

View Workflows with Security Issues

daily-copilot-token-report.lock.yml

• ⚠️ [Medium] artipacked (line 115) - Credential persistence through artifacts
• ℹ️ [Info] unverified_script_exec (lines 131, 143) - curl | sh for setup scripts

mcp-inspector.lock.yml

• ⚠️ [Low] template-injection (line 538) - Template expansion injection risk

ai-moderator.lock.yml

• ⚠️ [Warning] default_permissions_on_risky_events - Default permissions on risky events

copilot-setup-steps.yml

• ℹ️ [Info] unverified_script_exec (lines 17, 42) - curl | sh for uv and gh-aw installation

Other Workflows

All 149 compiled workflows contain:

• 📋 SC2129 style suggestions (grouped redirects)
• 📋 SC1003 info notices (single quote escaping)

---

Top Priority Issues

1. artipacked - Credential Persistence (Most Critical)

• Tool: zizmor
• Count: 1 occurrence
• Severity: Medium
• Affected: daily-copilot-token-report.lock.yml
• Why It Matters: If the workflow uploads artifacts containing credentials, they could be exposed
• Action Needed: Review if credential persistence in artifacts is necessary; if not, remove or sanitize

2. SC2129 - Shellcheck Style Issues (Most Common)

• Tool: actionlint/shellcheck
• Count: 164 occurrences (49% of all findings)
• Severity: Style (informational)
• Affected: All 149 workflows
• Why It Matters: Code readability and maintainability
• Root Cause: Compiler generates individual redirects instead of grouped redirects
• Action Needed: Update compiler to generate grouped command blocks

---

Fix Suggestion for SC2129 (Most Common Issue)

Issue: SC2129 - Consider using grouped redirects
Severity: Style (informational)
Affected Workflows: 164 occurrences across 149 workflows

Current Pattern (Generated by Compiler)

echo "line 1" >> "$GH_AW_PROMPT"
echo "line 2" >> "$GH_AW_PROMPT"
echo "line 3" >> "$GH_AW_PROMPT"

Recommended Pattern

{
  echo "line 1"
  echo "line 2"
  echo "line 3"
} >> "$GH_AW_PROMPT"

Prompt to Fix the Compiler

You are fixing the gh-aw workflow compiler to generate grouped redirects instead of individual redirects, eliminating shellcheck SC2129 warnings.

**Current Issue**:
The compiler generates multiple individual echo statements with separate `>>` redirects to the same file, triggering shellcheck SC2129 (style warning) in 164 locations across all compiled workflows.

**Required Fix**:
Modify the compiler to detect when multiple redirects target the same file and group them using brace syntax.

**Implementation Steps**:

1. **Locate Compiler Code**: Find where the compiler generates shell scripts with redirects (likely in prompt construction phase)

2. **Identify Pattern**: Look for code that generates multiple consecutive redirects to the same variable/file

3. **Implement Grouping Logic**:
   - Detect sequences of commands redirecting to the same file
   - Group them with `{ ... }` syntax
   - Use single redirect at the end

4. **Test**:
   - Recompile several workflows
   - Verify generated `.lock.yml` files use grouped redirects
   - Run `gh aw compile --actionlint` to confirm SC2129 warnings are eliminated
   - Test workflow execution to ensure behavior unchanged

5. **Edge Cases**:
   - Single redirects (don't group)
   - Mixed redirect types (`>` vs `>>`)
   - Redirects to different files

**Example Implementation**:

Before (current):
```bash
run: |
  echo "prompt_line_1" >> "$GH_AW_PROMPT"
  echo "prompt_line_2" >> "$GH_AW_PROMPT"
  echo "prompt_line_3" >> "$GH_AW_PROMPT"
```

After (desired):
```bash
run: |
  {
    echo "prompt_line_1"
    echo "prompt_line_2"
    echo "prompt_line_3"
  } >> "$GH_AW_PROMPT"
```

**Expected Impact**:
- Eliminates all 164 SC2129 warnings automatically on next compile
- Cleaner generated code
- Better performance (single file descriptor open)

**Testing**:
```bash
# Recompile and verify
gh aw compile --actionlint

# Count remaining SC2129 issues (should be 0)
gh aw compile --actionlint 2>&1 | grep SC2129 | wc -l
```

Please locate the relevant compiler code, implement the fix, and test with a few workflows.

Complete fix documentation: /tmp/gh-aw/cache-memory/fix-templates/actionlint-SC2129-fix.md

---

Historical Trends

Comparison with previous scan (February 10, 2026):

Metric	Feb 10	Feb 11	Change	Status
Total Findings	323	334	+11	⚠️ Slight increase
Zizmor (Security)	2	2	0	✅ Stable
Poutine (Supply Chain)	10	10	0	✅ Stable
Actionlint (Shellcheck)	320	322	+2	⚠️ Minor increase

Trend Analysis (7-day period)

• Feb 9: 336 findings (peak with 14 default_permissions warnings)
• Feb 10: 323 findings (significant improvement: 14 → 1 permission warnings)
• Feb 11: 334 findings (stable, +2 shellcheck issues)

Key Improvements:

• ✅ Supply chain security: Major reduction in default_permissions_on_risky_events warnings (14 → 1) between Feb 9 and Feb 10
• ✅ Security findings: Consistently 2 findings (stable and localized)

Areas to Watch:

• ⚠️ Shellcheck issues slightly increasing (320 → 322)
• ⚠️ 2 zizmor security findings remain unresolved since Feb 5

---

Recommendations

🔴 Immediate Priority (Security)

1. Fix artipacked issue in daily-copilot-token-report.lock.yml

   • Review if token persistence through artifacts is necessary
   • If needed, ensure artifacts are private and time-limited
   • If not needed, remove artifact upload or sanitize content
   • Impact: Prevents potential credential exposure

2. Review template-injection in mcp-inspector.lock.yml

   • Verify that template inputs are properly sanitized
   • Ensure no user-controlled input flows into template expansion
   • Impact: Mitigates code injection risk

🟡 Short-term (Code Quality)

3. Fix compiler to eliminate SC2129 warnings

   • Update compiler to generate grouped redirects
   • Test with subset of workflows first
   • Roll out to all workflows
   • Impact: Eliminates 164 style warnings, improves code quality
   • Effort: Medium (compiler modification + testing)

4. Address SC1003 shellcheck notices

   • Review if quote escaping changes are needed
   • May be acceptable as-is if generated code works correctly
   • Impact: Eliminates 158 info notices

🟢 Long-term (Process Improvement)

5. Evaluate curl | sh pattern alternatives

   • Consider pre-installed tools in runner images
   • Evaluate verified installation methods
   • Balance convenience vs. security best practices
   • Impact: Reduces supply chain risk notices

6. Add static analysis to CI/CD pipeline

   • Run zizmor/poutine/actionlint on PRs
   • Block merges if new critical/high issues introduced
   • Create automated tracking of findings over time
   • Impact: Prevents regression, maintains security posture

7. Document acceptable findings

   • Create exceptions list for intentional patterns
   • Document why certain findings are acceptable
   • Reduces noise in future scans

---

Next Steps

• Review and fix artipacked finding in daily-copilot-token-report workflow
• Investigate template-injection in mcp-inspector workflow
• Create issue/PR to update compiler for SC2129 fix
• Evaluate if SC1003 notices need addressing
• Consider adding static analysis to pre-commit or CI

---

Scan Metadata

• Scan Date: February 11, 2026
• Workflows Scanned: 149
• Compiled Workflows: 149 successful
• Compilation Warnings: 7 (action pinning, experimental features)
• Tools Version: actionlint 1.7.10, zizmor (latest), poutine (latest)
• Scan Data: Stored in /tmp/gh-aw/cache-memory/security-scans/2026-02-11.json
• Fix Templates: Available in /tmp/gh-aw/cache-memory/fix-templates/

Assessment: Overall security posture is good with no critical or high severity issues. The main findings are localized to 2 workflows and are well-understood. The majority (96.4%) of findings are style/informational issues from shellcheck that can be addressed through compiler improvements.

---

References:

• [Zizmor Documentation]((docs.zizmor.sh/redacted)
• Poutine Security Scanner
• Actionlint Documentation
• [ShellCheck SC2129]((www.shellcheck.net/redacted)
• [ShellCheck SC1003]((www.shellcheck.net/redacted)

AI generated by Static Analysis Report

• expires on Feb 18, 2026, 9:35 PM UTC

[github-actions[bot]]

🎉 The smoke test agent was here! All systems operational. 144 symbols found, builds passing, and ready to roll! 🚀

AI generated by Smoke Copilot

[github-actions[bot]]

🤖 Smoke test agent reporting in!

Just passed through running the full Copilot engine validation suite. All systems are green and the test harness is working beautifully! ✅

The static analysis report looks excellent - keeping security tight and code quality high. Keep up the great work!

Smoke test run: §21926186000

AI generated by Smoke Copilot

[github-actions[bot]]

💥 WHOOSH! 💥

🦸 The smoke test agent just swept through here at lightspeed!

BAM! ✅ All systems operational!
POW! 🔧 Tools verified!
KAPOW! 🚀 Workflow status: ACTIVE!

Up, up, and away! 🌟

---

Smoke Test Agent was here on 2026-02-11 23:46 UTC

AI generated by Smoke Claude

[pelikhan]

/plan

[github-actions[bot]]

🚀 Plan Command has started processing this discussion comment

[github-actions[bot]]

Plan Created ✅

I've analyzed the static analysis report and created 5 actionable sub-issues that address the key findings:

Security Issues (Priority)

1. Fix artipacked credential persistence - Medium severity finding in daily-copilot-token-report workflow
2. Review template-injection risk - Low severity finding in mcp-inspector workflow
3. Address default permissions - Supply chain warning in ai-moderator workflow

Code Quality & Process

4. Update compiler for grouped redirects - Eliminates 164 SC2129 style warnings across all workflows
5. Add static analysis to CI - Prevents regression by validating workflows in PRs

All issues are now grouped under a parent tracking issue and ready for assignment to GitHub Copilot agents. Each issue includes:

• Clear objectives and acceptance criteria
• Specific file locations and line numbers
• Detailed approach with step-by-step instructions
• Context from the static analysis findings

The plan prioritizes security fixes first, followed by systematic code quality improvements and process enhancements.