The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external input from specification files or feature descriptions without explicit boundary markers. This creates an attack surface where malicious instructions embedded in a spec file could attempt to influence the agent's behavior.
Ingestion points: The skill accepts a spec file path or description via the $ARGUMENTS variable in SKILL.md.
Boundary markers: There are no instructions to the agent to treat spec file content as untrusted or to ignore instructions contained within them.
Capability inventory: The agent has the authority to write files to the project directory and execute shell commands (go test, make).
Sanitization: No sanitization or validation of the input spec file content is described.
[COMMAND_EXECUTION]: The skill explicitly directs the agent to execute shell commands to verify code changes and run quality checks.
Evidence: Instructions in SKILL.md specify running make test-int, go test ./tests/integration, and make check.
[DYNAMIC_EXECUTION]: The workflow involves generating new Go source files and tests which are then executed by the agent as part of the TDD cycle.
Evidence: The skill defines patterns for creating command handlers (cmd/skillshare/), logic (internal/), and integration tests (tests/integration/) that are compiled and run during the 'RED' and 'GREEN' steps.

skillshare-implement-feature