The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill requires the agent to generate a local Node.js script (~/Desktop/Reels/analyse-reel.js) that incorporates the user-supplied reelUrl and executes it on the host machine to perform scraping and file system operations.
[DYNAMIC_CONTEXT_INJECTION]: The skill instructions in SKILL.md include shell command patterns (! export ...) which trigger automatic shell execution on specific platforms during the skill loading phase.
[PROMPT_INJECTION]: The instructions include specific commands such as "CRITICAL: Auto-start on load" and "Do not summarise," which are designed to override the agent's default operational procedures and output formatting.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources which could be used to influence the agent's behavior via the resulting script generation.
Ingestion points: Retrieval of Notion reference links (Step 1) and scraping of Instagram Reel content and transcripts (Step 3 and 4).
Boundary markers: The skill lacks explicit delimiters or warnings to ignore potentially malicious instructions embedded within the retrieved external content.
Capability inventory: The skill can write to the local filesystem (~/Desktop/Reels/), generate and execute Node.js scripts, and perform network operations via external APIs.
Sanitization: There is no evidence of sanitization or filtering of content retrieved from the Notion or Instagram links before it is analyzed by the AI model.
[COMMAND_EXECUTION]: The skill performs shell execution for environment variable configuration and runs the generated JavaScript files using the local Node.js runtime.
[EXTERNAL_DOWNLOADS]: The skill fetches data from Apify's Instagram scraping service and utilizes Google's Generative AI API for video content analysis.

reels-scripting