The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted markdown artifacts from the project directory to build a traceability matrix, creating a surface for indirect prompt injection. * Ingestion points: The skill reads vision.md, stakeholder_register.md, and all *.md artifacts within projects//// (defined in SKILL.md). * Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions within the processed data. * Capability inventory: The skill utilizes read, search, and editing capabilities, and references the execution of a python -m engine baseline command. * Sanitization: The skill does not implement content validation or sanitization for the ingested artifacts.
[DYNAMIC_EXECUTION]: The skill instructs the agent to execute a specific Python command for change control. * Evidence: references/scenario-spine-and-requirement-contract.md contains the instruction 'Baseline approved scope with python -m engine baseline snapshot'. This requires the presence of a specific local module.
[DATA_EXFILTRATION]: The skill documentation mentions the use of network checks in its fallback logic. * Evidence: SKILL.md specifies a fallback behavior 'if a required file, reviewer, standard source, network check, renderer, or execution capability is unavailable'.

09-traceability-engineering