sdlc-write-spec
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from various wiki files which could contain malicious instructions.\n
- Ingestion points: The agent reads wiki/index.md, wiki//features/index.md, overview.md, log.md, prd.md, and .wiki-llm/schema/*.md (files referenced in SKILL.md and references/read-protocol.md).\n
- Boundary markers: No explicit delimiters or instructions to ignore embedded prompts are used when reading external wiki content.\n
- Capability inventory: The skill can execute a local Python script (.wiki-llm/tools/wiki_llm.py), perform git commands (fetch, show, grep, commit, worktree), and write files to the repository.\n
- Sanitization: No sanitization or validation of the content read from the wiki is performed before it is used in the agent's context.\n- [COMMAND_EXECUTION]: The skill is configured to execute a Python script residing within the target repository.\n
- Evidence: In references/read-protocol.md, the tool is defined as T="python3 .wiki-llm/tools/wiki_llm.py". This script is executed to perform wiki checks and management tasks as described in SKILL.md and references/write-protocol.md.
Audit Metadata