he-spec

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell scripts and git commands to manage the specification lifecycle. \n
  • Evidence: It runs bash scripts/runbooks/select-runbooks.sh --skill he-spec (SKILL.md) to dynamically load project-specific instructions. \n
  • Evidence: It performs git operations including add, commit, and diff (SKILL.md) to track changes in the docs/specs/ directory. \n- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it is designed to ingest and parse untrusted external documents. \n
  • Ingestion points: The "Phase 0-pre: Detect External Input" section in SKILL.md instructs the agent to read and parse external specifications, PRDs, or requirements documents provided via file paths, pasted content, or URLs. \n
  • Boundary markers: There are no instructions to wrap external content in delimiters or to include "ignore embedded instructions" warnings when processing this data. \n
  • Capability inventory: The skill has the capability to write files to the docs/specs/ directory, execute git commands, and run a local bash script. \n
  • Sanitization: No sanitization, escaping, or validation of the external content is performed before it is mapped into the final specification artifact.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 07:04 AM
Security Audit — agent-trust-hub — he-spec