searching-sourcegraph
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (code, commits, file contents) from Sourcegraph-indexed repositories, creating a surface for malicious instructions to influence the agent.
- Ingestion points: Content retrieved via
read_file,deepsearch_read,keyword_search,nls_search, andcommit_searchtools. - Boundary markers: Absent; the instructions do not specify any delimiters or warnings to ignore embedded instructions within retrieved source code or comments.
- Capability inventory: The agent can read files, trace symbols, list repositories, and synthesize code or documentation based on findings.
- Sanitization: Absent; there is no validation or filtering of content fetched from remote repositories before it is processed by the agent.
- [DATA_EXFILTRATION]: The skill includes guidance and patterns for locating sensitive configuration files, which increases the risk of accidental exposure of secrets.
- Evidence: In
query-patterns.md, the skill provides a specific regex pattern (file:\.env\. [A-Z_]+=) to help the agent find environment configuration files, which typically contain sensitive credentials or API keys.
Audit Metadata