The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze potentially untrusted data, including code diffs, pull request descriptions, and project documentation.
Ingestion points: The agent is instructed to read the active checkout's diffs, along with files such as AGENTS.md, CONTRIBUTING.md, and docs/architecture.md (SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings for content within the diffs, instead directing the agent to treat these repository documents as authoritative sources of truth.
Capability inventory: The skill instructions suggest the agent has the capability to execute shell commands via the just task runner to verify project state (e.g., just check, just db-migrate).
Sanitization: There is no evidence of sanitization or filtering of external content before it is processed by the agent.
[COMMAND_EXECUTION]: The skill directs the agent to perform various shell-based operations using the local project's task runner.
Evidence: The instructions include calls to just db-reset-local, just test-package, just check, just graphql-codegen, and just commit-check across multiple sections of SKILL.md.
Context: These commands are used for standard development and verification workflows within the project's ecosystem.

code-review-guardrails