linkedin-message-writer
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external LinkedIn profiles and posts which is then used to generate personalized messages.
- Ingestion points: Profile data (headlines, work history, skills) and recent post content fetched via Apify actors (SKILL.md Phase 2).
- Boundary markers: Absent. There are no explicit instructions to use delimiters or "ignore embedded instructions" guards for the data retrieved from the scraper.
- Capability inventory: The skill uses
curlfor network requests to Apify and performs file-write operations to export CSV campaign files. - Sanitization: Absent. The skill generates messages based directly on the raw strings returned by the API.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to interact with external APIs.
- Evidence: Phase 2 (Research) provides explicit
curlcommands for the agent to run to trigger and poll Apify actors. - [EXTERNAL_DOWNLOADS]: The skill fetches data from Apify, an external web scraping service.
- Evidence: Requests are made to the
api.apify.comdomain to retrieve lead intelligence.
Audit Metadata