The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches bucket names, configurations, and file metadata from external Supabase API endpoints. This data is ingested into the agent's context to generate reports and update tracking files, creating a potential surface for indirect prompt injection.
Ingestion points: Data retrieved from Supabase Storage API endpoints (/storage/v1/bucket, /storage/v1/object/list/[bucket]).
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potentially malicious instructions embedded in bucket names or metadata.
Capability inventory: The skill performs local file writes (.sb-pentest-context.json, .sb-pentest-audit.log) and executes network operations via curl.
Sanitization: There is no evidence of sanitization or validation of the data received from the external API before it is processed or written to files.
[EXTERNAL_DOWNLOADS]: The skill interacts with the Supabase Storage API to retrieve configuration data. Supabase is a well-known service, and these network requests are consistent with the skill's stated purpose of auditing project storage.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using curl to interact with the Supabase API and requires the agent to progressively write findings to local filesystem paths.

supabase-audit-buckets-list