The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes node scripts/run-ast-analysis.mjs analyze . to perform Abstract Syntax Tree (AST) analysis on the local project directory for deep code inspection.- [COMMAND_EXECUTION]: Provides shell commands for users to monitor progress by running cat .stackshift-state.json | jq in the terminal.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted source code from the local directory to drive the 'Reverse Engineer' and 'Create Specifications' phases.
Ingestion points: The local file system is scanned by the run-ast-analysis.mjs script and the stackshift:code-analyzer agent.
Boundary markers: The skill does not define explicit boundary markers or instructions for the agent to ignore potentially malicious content within the ingested code comments or structures.
Capability inventory: The skill possesses significant capabilities, including reading the entire codebase, writing generated documentation to docs/reverse-engineering/, creating feature specifications in .specify/, and performing code implementation in 'Gear 6'.
Sanitization: No sanitization or validation of the ingested code content is mentioned before it is processed by the LLM to generate implementation plans.

cruise-control