phase-gated-debugging
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation references an external repository at
https://github.com/krabat-l/claude-debug. This is provided as a methodology reference and source attribution; it is not bundled as a dependency and does not involve automated script execution. - [INDIRECT_PROMPT_INJECTION]: The skill is inherently designed to process external, untrusted content such as source code, logs, and error messages to perform its primary function of debugging.
- Ingestion points: External source code and command/test output logs in Phases 1 (Reproduce) and 2 (Isolate).
- Boundary markers: The instructions do NOT specify delimiters or "ignore embedded instructions" warnings for the code context being analyzed.
- Capability inventory: The skill manages file reads, reproduction command/test executions, diagnostic logging, and code modification (Phase 4).
- Sanitization: There are no explicit sanitization or validation logic steps for the external data being processed.
- [NO_CODE]: The skill consists primarily of process instructions and natural language guidelines without including its own scripts, binaries, or configuration files.
Audit Metadata