dx-devops-test-pipeline-configure

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the sf CLI to interact with Salesforce orgs, performing both read and write operations.
  • It executes sf data query to retrieve pipeline, stage, and trigger information from the org.
  • It makes authenticated REST API calls using sf api request rest to Salesforce Connect API endpoints (e.g., /services/data/v67.0/connect/devopstesting/qualityGate).
  • It creates and updates records directly using sf data create record and sf data update record for rule configuration and linking.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes metadata and records retrieved from the Salesforce org, which serves as a potential vector for indirect prompt injection.
  • Ingestion points: Data retrieved from standard objects such as DevopsPipeline, DevopsTestSuite, and DevopsPipelineStage via CLI queries.
  • Boundary markers: The skill requires mandatory impact previews and explicit confirmation gates (e.g., in SKILL.md Step 3 and references/configuring-quality-gate.md) before any write operations are executed.
  • Capability inventory: The agent can create records, update existing records, and trigger Connect API actions within the Salesforce environment.
  • Sanitization: There is no explicit mention of sanitization for org-retrieved values (like suite names or provider IDs) before they are used to build CLI commands or API payloads, relying instead on the structured nature of the sf commands and user oversight.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 03:37 AM
Security Audit — agent-trust-hub — dx-devops-test-pipeline-configure