Validate agent handoff packets and resume readiness using schema, freshness, and replay checks. Use when tasks pause/resume across sessions, agents, or humans — including when a user wants to continue where they left off, hand off to another agent, resume a previous task, or pick up an interrupted workflow. Includes explicit untrusted-content/prompt-injection guardrails for third-party inputs.
96
Quality
100%
Does it follow best practices?
Impact
96%
1.50xAverage score across 9 eval scenarios
At the end of each quarter, an engineering manager reviews all open handoff packets left by agents and engineers across active projects. The goal is to identify which tasks are in a state that can be resumed immediately, which ones require some remediation before resuming, and which ones represent a more serious problem that needs escalation.
Three handoffs from the current quarter are ready for audit. They are from different projects and were prepared by different team members. The manager needs a structured report that gives an at-a-glance status for each handoff and enough detail to take action.
Write the audit report to quarterly-audit.md. For each handoff, provide enough detail that the manager can act on the findings immediately. Conclude with a summary that shows the overall status of all three projects at a glance.
The following files are provided as inputs. Extract them before beginning.
=============== FILE: handoff-alpha.json =============== { "objective": "Implement full-text search for the product catalog using Elasticsearch", "completed": [ "Set up Elasticsearch 8.x cluster on staging", "Indexed all 50,000 product records", "Implemented search API endpoint with pagination" ], "unresolved": [ "Performance under 1000 concurrent queries not yet benchmarked" ], "assumptions": [ "Elasticsearch cluster will be scaled to 3 nodes before production launch" ], "next_action": "Run load test with k6 at 1000 concurrent users and document results", "risks": [ "Index mapping changes require full reindex — plan migration window carefully" ], "updated_at": "2025-12-02T09:00:00Z", "resume_token": "es-search-impl-prod-4nBk7wX2" }
=============== FILE: handoff-beta.json =============== { "objective": "Migrate CI/CD pipeline from Jenkins to GitHub Actions", "completed": [ "Documented all existing Jenkins jobs", "Converted 8 of 22 pipelines to GitHub Actions workflows" ], "unresolved": [ "3 pipelines use a custom Jenkins plugin with no GitHub Actions equivalent" ], "assumptions": [ "DevOps team will handle the 3 custom plugin pipelines separately" ], "next_action": "Continue converting remaining 11 standard pipelines", "risks": [ "Build times may increase with GitHub Actions runner cold starts" ], "updated_at": "2024-01-10T14:30:00Z", "resume_token": "cicd-gh-migration-v2-mN5pQr8j" }
=============== FILE: handoff-gamma.json =============== { "objective": "Build automated alerting for database connection pool exhaustion", "completed": [ "Defined alert thresholds with the SRE team" ], "unresolved": [ "Prometheus scrape config for the connection pool metrics endpoint not yet added" ], "assumptions": [], "next_action": "", "risks": [ "Without alerting, pool exhaustion will only be detected after user-facing errors" ], "updated_at": "2025-11-30T16:45:00Z", "resume_token": "db-alert-setup-sre-7hJk3Lm9" }