The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses a local file at ~/.config/gong/credentials.json to retrieve Gong API credentials (access_key, secret_key, and base_url). This is the primary mechanism for authenticating requests to the Gong service.
[PROMPT_INJECTION]: The skill retrieves call transcripts and meeting content, which represents untrusted external data.
Ingestion points: The transcript and call commands in scripts/gong.sh fetch raw text from the Gong API.
Boundary markers: There are no markers or instructions provided to the agent to distinguish between the fetched transcript text and system instructions.
Capability inventory: The script uses curl for network requests and jq for parsing JSON data.
Sanitization: The skill does not perform sanitization or filtering on the retrieved transcript content before it is presented to the agent.

gong