The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by design, as it requires the agent to process untrusted data from external sources.
Ingestion points: The agent is instructed to read and respond to customer reviews (Phase 4) and monitor competitor reviews (Phase 5) from platforms like G2 and TrustRadius, as described in SKILL.md and references/trustradius-playbook.md.
Capability inventory: The skill includes scripts/check-output.py which executes local file system read operations to validate deliverables.
Boundary markers: There are no instructions or delimiters provided to ensure the agent ignores potentially malicious instructions embedded in the external review text.
Sanitization: The skill lacks logic to sanitize or filter external content before processing it for response generation or intelligence reporting.

review-platforms