merge-reconciler

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the terminal tool to execute standard git operations including git status, git merge-base, git log, git diff, git add, and git commit. It also instructs the agent to run the project's own build and test commands to verify the merge result.- [PROMPT_INJECTION]: The skill is designed to process untrusted data from git history and task management systems.
  • Ingestion points: Reads external data from git log messages, git diff content, PR bodies, and hermes kanban task summaries.
  • Boundary markers: Instructs the agent to re-state intents in its own words before proceeding, creating a conceptual boundary, though no formal escape delimiters are used for the raw diff content.
  • Capability inventory: Includes shell command execution (terminal) and file system writes (patch, write_file).
  • Sanitization: No specific sanitization or filtering is applied to the ingested text, which is typical for a code reconciliation tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 01:59 PM
Security Audit — agent-trust-hub — merge-reconciler