Self-maintaining codebase memory for AI coding agents: an AST index that answers which files matter, plus agent-written notes that flag themselves stale when the code changes. No embeddings, no API key. CLI + MCP.
The publisher performed an authenticated act of publication — proving control of a namespace in the official MCP registry, or appearing in a first-party catalogue an operator curates. That establishes WHO published it. It is not a security finding, and attested artifacts carry Flagged verdicts too.
253 critical findings reported by cisco-skill-scanner, agent-audit-kit, 8 of them in the installed artifact. The composite score alone does not reflect this — read the per-scanner reports below.
This artifact ships from 2 sources and their verdicts differ. The worst is github (Flagged). Installing from a cleaner source does not mean the others are safe — check the source you actually install from.
github: https://github.com/AkashGoenka/coldstart, 624 findings
Choose your runtime: the CLI writes the mcpServers config entry for that target.
nerlo install --target claude-code -- @cstart/coldstart
Install respects the composite badge: Clean proceeds, Caution prompts for confirmation, and Flagged is refused. Write operations require an API token.
6 of 11 scanners examined this package. The rest do not apply to it — language, ecosystem and packaging decide which analyzers can say anything, and a scanner that cannot examine an artifact reports nothing rather than passing it.
v1.4.0
6 findings · 3.2s
v2.0.11
318 findings · 73.4s
v0.3.26
246 findings · 34.4s
v0.1.0
0 findings · 61.7s
v0.1.0
0 findings · 61.7s
v0.1.0
0 findings · 25.3s
No report in this scan: capslock, govulncheck, osv-scanner, trivy, trivy_image.
3 scans on record. Every scan's full results are retained immutably for 24 months.
| Completed | Composite | Change | Scanners | Status |
|---|---|---|---|---|
| Aug 3, 2026 | Scan Halted | — | agentshield: completecisco-skill-scanner: completeagent-audit-kit: completenerlo-behavioral: completenerlo-install-instruction: completenerlo-multi-source: complete | rejected |
| Jul 30, 2026 | Scan Halted | — | cisco-skill-scanner: completeagent-audit-kit: completenerlo-behavioral: completenerlo-install-instruction: completenerlo-multi-source: complete | rejected |
| Jul 30, 2026 | Scan Halted | — | cisco-skill-scanner: completeagent-audit-kit: completenerlo-behavioral: completenerlo-install-instruction: completenerlo-multi-source: complete | rejected |
Findings in files this artifact installs stand as reported; a model re-read the rest and said which ones it believes are false positives. This is a second opinion published beside the evidence, not a correction to it: the per-scanner reports above are unchanged, every dismissed finding is still listed there at its original severity, and the score and the Flagged badge are computed from those raw severities alone. Nothing below moved them.
Out of 570 findings across every scanner that examined this package. 245 of 253 findings rated critical by their scanner were among the dismissals.
Two keys had to turn. The model had to name a reason from a fixed list of six, and a deterministic check of ours — over which files this artifact actually installs — had to independently agree. Anything it could not corroborate is in “not reviewed” above, at full severity.
234 Agent instruction file — The match is in a file that instructs an AI coding agent — a CLAUDE.md or similar. It is documentation of intent, and our own check confirmed it is not part of what gets installed.
Rules: AAK-AGENT-001, AAK-AGENT-003 · Files: CLAUDE.md · reported by agent-audit-kit
227 Test or fixture file — The match is in a test or fixture. Our own check confirmed the file is not part of what gets installed.
Rules: COMMAND_INJECTION_JS_CHILD_PROCESS, AAK-LEGAL-002, FILE_MAGIC_MISMATCH and 1 more · Files: /repo/tests/freshness-b3b7.test.ts, tests/fixtures/npm-workspaces/package.json, tests/fixtures/npm-workspaces/packages/shared-utils/package.json and 25 more · reported by cisco-skill-scanner, agent-audit-kit
2 Documentation or example — The match is in documentation or sample code, confirmed by our own check to be outside the installed artifact.
Rules: SOCIAL_ENG_VAGUE_DESCRIPTION, MANIFEST_MISSING_LICENSE · Files: /repo/docs/SKILL.md · reported by cisco-skill-scanner
These findings are in files this artifact installs and are not placeholder values. No false-positive basis can be corroborated for such a finding, so this review never dismisses one; they were not sent to the model and stand at their scanner's severity.
47 DATA_EXFIL_JS_FS_ACCESS
Files: /repo/src/init.ts, /repo/src/migrate.ts, /repo/src/unwire.ts · reported by cisco-skill-scanner
15 BINARY_FILE_DETECTED
Files: /repo/vendor/wasm/tree-sitter-c-sharp.wasm, /repo/vendor/wasm/tree-sitter-typescript.wasm, /repo/vendor/wasm/tree-sitter-cpp.wasm and 12 more · reported by cisco-skill-scanner
15 UNANALYZABLE_BINARY
Files: /repo/vendor/wasm/tree-sitter-yaml.wasm, /repo/vendor/wasm/tree-sitter-c-sharp.wasm, /repo/vendor/wasm/tree-sitter-java.wasm and 12 more · reported by cisco-skill-scanner
4 AAK-HOOK-RCE-001
Files: hooks/recall-query.mjs, hooks/recall-seen.mjs, hooks/cursor-kb-recall.mjs and 1 more · reported by agent-audit-kit
4 COMMAND_INJECTION_JS_CHILD_PROCESS
Files: /repo/scripts/bench-find.mjs, /repo/scripts/probe-repos.mjs, /repo/scripts/wasm-probe-compare.mjs · reported by cisco-skill-scanner
3 FILE_MAGIC_MISMATCH
Files: /repo/src/unwire.ts, /repo/site/llms.txt, /repo/src/kb/view-template.ts · reported by cisco-skill-scanner
2 AAK-HEALTHCARE-AI-004
Files: CHANGELOG.md, TROUBLESHOOTING.md · reported by agent-audit-kit
2 MANIFEST_MISSING_LICENSE
Files: /repo/SKILL.md, /repo/templates/SKILL.md · reported by cisco-skill-scanner
2 SOCIAL_ENG_VAGUE_DESCRIPTION
Files: /repo/SKILL.md, /repo/templates/SKILL.md · reported by cisco-skill-scanner
1 AAK-ANTHROPIC-SDK-001
Files: src/server/mcp.ts · reported by agent-audit-kit
1 AAK-OAUTH-3P-001
Files: package.json · reported by agent-audit-kit
1 AAK-SEC-MD-001
Files: SECURITY.md · reported by agent-audit-kit
1 AAK-SUPPLY-005
Files: package-lock.json · reported by agent-audit-kit
1 EXCESSIVE_FILE_COUNT
Files: /repo/. · reported by cisco-skill-scanner
1 hooks-code-transcript-access-112
Files: hooks/cursor-kb-elicit.mjs · reported by agentshield
1 hooks-code-transcript-access-261
Files: hooks/kb-elicit.mjs · reported by agentshield
1 hooks-code-transcript-access-89
Files: hooks/codex-kb-recall.mjs · reported by agentshield
1 hooks-code-transcript-access-90
Files: hooks/cursor-kb-recall.mjs · reported by agentshield
1 hooks-code-transcript-access-93
Files: hooks/codex-kb-elicit.mjs · reported by agentshield
1 hooks-code-transcript-access-98
Files: hooks/kb-recall.mjs · reported by agentshield
1 LOW_ANALYZABILITY
Files: /repo · reported by cisco-skill-scanner
1 OVERSIZED_FILE
Files: /repo/vendor/wasm/tree-sitter-c-sharp.wasm · reported by cisco-skill-scanner
Critical RCE and command injection vulnerabilities identified.
The artifact exhibits critical remote code execution capabilities in multiple hook files, including `hooks/codex-kb-elicit.mjs` and `hooks/recall-seen.mjs`. It also contains critical command injection vulnerabilities through child process execution in scripts such as `/repo/scripts/bench-find.mjs` and `/repo/scripts/probe-repos.mjs`. Additionally, a high severity finding indicates potential credential access or sensitive data exposure related to an Anthropic SDK in `src/server/mcp.ts`. These findings collectively indicate significant risks related to shell execution and potential data compromise.
Concern level assessed by the review: Critical. This is the review's scale, not the registry badge.
Reviewed by gemini-2.5-flash. The model sees the findings, not your code, and cannot change a score.