dd-apm
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs users to install the
puptool usingbrew tap datadog-labs/pack. This is a standard installation method for vendor-provided CLI tools and points to the official Datadog Labs repository. - [COMMAND_EXECUTION]: The skill makes extensive use of the
pupcommand to perform authenticated operations against the Datadog API, including querying production telemetry data and managing SLO configurations. These commands are consistent with the skill's stated purpose of performance analysis. - [PROMPT_INJECTION]: The skill processes data from application traces and logs, which is a potential surface for indirect prompt injection.
- Ingestion points: External data enters the context through
pup traces searchandpup traces aggregatecommands used inSKILL.md. - Boundary markers: No specific delimiters are used to separate CLI output from agent instructions.
- Capability inventory: Subprocess execution is limited to the
puputility for data retrieval and resource creation. - Sanitization: The skill assumes standard output from the CLI tool; sanitization is handled by the underlying agent framework and the tool itself.
Audit Metadata