The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted project data (source code, API schemas, and existing tests) to generate BDD scenarios and autotests.\n
Ingestion points: Target project analysis phases in mode-install.md (S2) and mode-create.md (C2).\n
Boundary markers: The skill uses a structured delegation contract (GOAL, ROLE, SCOPE, CONTEXT, CONSUMER, DONE) to isolate sub-agent tasks.\n
Capability inventory: Access to Bash, Write, Edit, and Agent tools across various scripts.\n
Sanitization: Key decisions and artifacts are validated through AskUserQuestion and multi-agent quorum reviews.\n- [COMMAND_EXECUTION]: The skill executes local shell scripts for environment detection and triggers project-specific build tools to validate generated tests.\n
Evidence: detect-mode.sh is executed in SKILL.md to resolve versions and modes.\n
Evidence: Build commands like mvn compile, npx tsc, and dotnet build are utilized in mode-create.md for compilation checks.

e2e