The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The generated statistics page template fetches the Chart.js library from cdn.jsdelivr.net. This is a well-known, high-reputation content delivery network used to provide visualization capabilities to the research assets created by the skill.
[PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it ingests and processes untrusted journalist requests from the Qwoted platform. A malicious request could attempt to influence the agent's drafting behavior. The skill mitigates this risk by:
Ingestion points: Journalist requests are fetched via qwoted_search.py and analyzed in SKILL.md.
Boundary markers: SKILL.md mandates that all pitches must be shown to the user for explicit approval before they are sent.
Capability inventory: The agent can execute local Python scripts, perform web research, and write HTML files to the local filesystem.
Sanitization: The skill relies on the user's manual review of the generated pitches and statistics pages to ensure content integrity.
[COMMAND_EXECUTION]: The skill executes local Python scripts (qwoted_login.py, qwoted_search.py, qwoted_pitch.py, qwoted_profile.py) and the playwright CLI to automate browser interactions. These executions are confined to the intended functionality of interacting with the Qwoted platform and are triggered by the agent within the local environment.
[SAFE]: The skill manages sensitive Qwoted session cookies by storing them in a local directory (~/.qwoted/). The documentation and code include appropriate warnings and measures (like .gitignore recommendations) to ensure these credentials remain private and are not committed to version control.

qwoted-seo-backlinks