tlc-spec-driven

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local command-line tools for project maintenance and analysis. Specifically, it uses git for atomic commits in the implementation phase and tools like grep, ripgrep (rg), and ast-grep (sg) for codebase exploration and mapping.\n- [EXTERNAL_DOWNLOADS]: The skill identifies installation via the @tech-leads-club/agent-skills package and recommends the use of ast-grep, a well-known structural search tool, for precise code analysis.\n- [PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection, which is common in tools designed to analyze external data or codebases.\n
  • Ingestion points: The agent ingests data from local project files during brownfield mapping (references/brownfield-mapping.md) and from external web search results during the research phase (references/design.md).\n
  • Boundary markers: There are no explicit instructions or delimiters defined to separate untrusted data from the agent's system instructions, which could lead the agent to inadvertently follow instructions embedded in analyzed files.\n
  • Capability inventory: The skill allows the agent to modify the filesystem, execute search commands, and perform git operations (references/implement.md, references/code-analysis.md).\n
  • Sanitization: No specific sanitization, validation, or filtering mechanisms are described for handling data retrieved from the filesystem or external sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 03:56 PM