The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several shell and version control commands to perform its duties.
Evidence includes git ls-files, find, git rev-parse, git diff, and git checkout for scope enumeration, backup creation, and fix verification.
[DYNAMIC_EXECUTION]: The skill generates and executes reproducer scripts to verify logic bugs before applying fixes.
Phase 0a in guide-phases-0-2-consent-scope-health.md specifies that these scripts are intended for pure computation with strict limits: no network access, no file writes, no database access, and a 10-second timeout.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository files (source code, config, documentation) and has the capability to write to the filesystem and execute scripts, creating a potential surface for indirect injection.
Ingestion points: Files throughout the repository are scanned during the health and deep review phases (Phase 1 and 3).
Boundary markers: The skill uses a semi-formal tracing methodology (Premises → Trace → Divergence) which provides a structured approach, but it does not explicitly define prompt boundary markers to separate data from instructions.
Capability inventory: The skill can write to source files, execute git commands, and run generated reproducer scripts.
Sanitization: The skill does not explicitly detail sanitization or escaping of the ingested file content before it is processed by the logic review phase.

logic-fix-all