estimate-actual

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Uses standard utilities like git and gh (GitHub CLI) to retrieve diffs, logs, and pull request information for estimation analysis.\n- [COMMAND_EXECUTION]: Executes a local script .claude/scripts/code_graph to perform dependency tracing and calculate the blast radius of changes.\n- [DATA_EXPOSURE]: Accesses local project documentation and configuration files to identify relevant patterns and settings.\n- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external data including git diffs and project plans.\n
  • Ingestion points: Reads git diff, git log, gh pr view, and plan markdown files.\n
  • Boundary markers: The skill instructions do not define explicit delimiters for ingested external content.\n
  • Capability inventory: Includes shell execution for git/gh tools and file-write access to plans/_estimation-samples.csv.\n
  • Sanitization: No explicit sanitization or filtering of ingested data is mentioned in the protocol.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 03:11 PM
Security Audit — agent-trust-hub — estimate-actual