The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection by processing external data from multiple sources. \n- Ingestion points: The skill is instructed in SKILL.md to load venture context from a local file named innov.local.md and to process user-provided pilot metrics and customer feedback. \n- Boundary markers: Absent; there are no specific delimiters or instructions provided to the model to differentiate between data and instructions within the ingested content. \n- Capability inventory: The skill is limited to providing textual business analysis and does not have access to high-risk tools like shell execution, network access, or file system writes. \n- Sanitization: Absent; the skill does not perform any sanitization, escaping, or validation on the data it processes. \n- [SAFE]: No malicious patterns such as direct prompt injection, credential exfiltration, or obfuscation were detected. The skill logic is consistent with its stated purpose and includes appropriate instructional guardrails to prevent analytical errors and correctly redirects out-of-scope tasks.

validate