Skip to content

richfrem/agent-plugins-skills / plugins/agent-orchestration/.apm/skills/orchestrating-agent-work

orchestrating-agent-work

Verified

Interactive orchestrator for complex engineering work. Analyzes the task, conducts a diagnostic interview to select the appropriate execution strategy (Directed Acyclic Graph versus Loop, and which Loop type), plans the artifact, and supervises execution and verification. Use for "how should I orchestrate this", "plan this workflow", "should this be a DAG or a loop", "run these jobs in parallel", "delegate this to a worker agent", or "run/resume this graph manifest". Not for single trivial edits or skill self-improvement.

FileKindSize
SKILL.mdskill md6.8 KB
assets/templates/dual-loop-meta-tasks.mdmarkdown1.6 KB
assets/templates/loop_retrospective_template.mdmarkdown1.5 KB
assets/templates/red_team_briefing_template.mdmarkdown747 B
assets/templates/scratchpad-template.mdmarkdown991 B
assets/templates/strategy-packet-template.mdmarkdown1.2 KB
assets/templates/workflow-retrospective-template.mdmarkdown1.3 KB
evals/evals.jsonconfig7.0 KB
references/PATTERN_GUIDE.mdmarkdown6.2 KB
references/acceptance-criteria.mdmarkdown6.8 KB
references/agentic-os-reuse-map.mdmarkdown2.5 KB
references/bot-delegation-patterns.mdmarkdown5.2 KB
references/cheapest_models.jsonconfig845 B
references/cheapest_models.mdmarkdown1.1 KB
references/cli-agent-executor.mdmarkdown3.5 KB
references/delegated-dag-refactoring-playbook.mdmarkdown5.5 KB
references/diagrams/agent_swarm.mmdother1.8 KB
references/diagrams/dual_loop_architecture.mmdother1.4 KB
references/diagrams/graph-planner-flow.mmdother540 B
references/diagrams/graph_execution_state_machine.mmdother823 B
references/diagrams/orchestrating_agent_work_architecture.mmdother4.5 KB
references/dispatch-preflight.mdmarkdown2.1 KB
references/fallback-tree.mdmarkdown9.7 KB
references/graph-execution.mdmarkdown4.1 KB
references/graph-manifest-schema.jsonconfig7.0 KB
references/graph-planning.mdmarkdown3.0 KB
references/learning-loop.mdmarkdown4.9 KB
references/patterns.jsonconfig2.0 KB
references/post_run_survey.mdmarkdown1.4 KB
references/routing-and-closure.mdmarkdown2.4 KB
references/strategy-agent-swarm.mdmarkdown2.1 KB
references/strategy-dual-loop.mdmarkdown1.6 KB
references/strategy-red-team-review.mdmarkdown2.0 KB
references/strategy-selection.mdmarkdown4.0 KB
references/topology-rules.mdmarkdown6.0 KB
scripts/agent_orchestrator.pyscript7.6 KB
scripts/dispatch_guard.pyscript1.1 KB
scripts/graph_runner.pyexecutablescript36.3 KB
scripts/select_strategy.pyscript13.1 KB
scripts/swarm_run.pyscript28.0 KB
scripts/validate_manifest.pyexecutablescript19.8 KB
scripts/work_loop.pyscript28.0 KB
tests/_loop_fixtures.pyscript2.5 KB
tests/test_graph_fixes.pyscript12.6 KB
tests/test_graph_runner.pyscript30.0 KB
tests/test_isolation.pyscript3.5 KB
tests/test_loop_strategies.pyscript7.7 KB
tests/test_select_strategy.pyscript7.8 KB
tests/test_swarm_run.pyscript9.4 KB
tests/test_swarm_run_models.pyscript3.1 KB
tests/test_validate_manifest.pyscript18.1 KB
tests/test_work_loop_e2e.pyscript3.8 KB
tests/test_work_loop_hardening.pyscript27.3 KB

references/diagrams/orchestrating_agent_work_architecture.mmd

View on GitHub
---
title: Orchestrating Agent Work - Unified Architecture
---
flowchart TD
    subgraph Phase1["1. Interactive Triage & Strategy Selection"]
        direction TB
        Task["Incoming Engineering Task<BR>(Task Brief / Prompt / Requirements)"]
        Analyze["Analyze Task Structure<BR>(Dependencies, Ordering Barriers, Concurrency, Risk)"]
        Interview["Interactive Interview (1-3 Questions)<BR>(Unit structure, Ordering/convergence, Assurance)"]
        SelectScript["Run scripts/select_strategy.py<BR>(Output: select-loop-strategy-decision.json)"]
        Confirm["Human Confirmation / Override<BR>(Confirm Graph vs. Loop before planning)"]

        Task --> Analyze
        Analyze --> Interview
        Interview --> SelectScript
        SelectScript --> Confirm
    end

    Confirm -->|"Multiple steps, parallel fan-out,<BR>join barriers, or ordered mutations"| GraphBranch["Strategy: Graph (DAG)"]
    Confirm -->|"Single bounded change<BR>(Supervisor / Worker)"| DualLoopBranch["Strategy: Dual-Loop"]
    Confirm -->|"Many identical independent tasks<BR>(Bulk batch operations)"| SwarmBranch["Strategy: Parallel Swarm"]
    Confirm -->|"Minimal localized edit"| DirectBranch["Strategy: Direct Edit"]

    subgraph Phase2["2. Plan"]
        direction TB
        subgraph GraphPlan["Graph Planning (graph-planning.md)"]
            DAGAuthor["Architect Nodes & Dependencies<BR>(Read fan-out, Join barriers, Verifier gates)"]
            DAGValidate["Validate Manifest<BR>(scripts/validate_manifest.py)"]
            DAGManifest["Emit graph-manifest.json"]
            DAGAuthor --> DAGValidate --> DAGManifest
        end

        subgraph DualLoopPlan["Dual-Loop Planning (strategy-dual-loop.md)"]
            TaskPacket["Author Task Packet<BR>(handoffs/task_packet_NNN.md)<BR>Enforce 'NO GIT' worker rule"]
        end

        subgraph SwarmPlan["Swarm Planning (strategy-agent-swarm.md)"]
            SwarmJob["Partition Tasks & Author Job<BR>(docs/plans/work-tasks/<id>.job.md)<BR>Build file checklist"]
        end

        DirectPlan["Author Implementation Plan"]
    end

    GraphBranch --> GraphPlan
    DualLoopBranch --> DualLoopPlan
    SwarmBranch --> SwarmPlan
    DirectBranch --> DirectPlan

    subgraph Phase3["3. Execute"]
        direction TB
        subgraph GraphExec["Graph Execution (graph-execution.md)"]
            GraphRunner["Deterministic DAG Runner<BR>(scripts/graph_runner.py)"]
            WorktreeIso["Isolated Git Worktree<BR>(Mutations confined to .worktrees/)"]
            StateReceipts["Per-Node Commits & Receipts<BR>(survives rollback outside worktree)"]
            Rollback["Automatic Rollback on Failure<BR>(Restores clean start SHA)"]
            GraphRunner --> WorktreeIso --> StateReceipts
            StateReceipts -.->|Failure| Rollback
        end

        subgraph DualLoopExec["Dual-Loop Execution"]
            WorkerDispatch["Dispatch Inner Worker<BR>(CLI dispatch with stdin < /dev/null)"]
            NoGitWorker["Worker writes code & runs tests<BR>(Forbidden from running git write commands)"]
            WorkerDispatch --> NoGitWorker
        end

        subgraph SwarmExec["Swarm Execution"]
            SwarmRun["Parallel Worker Dispatch<BR>(scripts/swarm_run.py --workers N)"]
            BatchResume["Independent worker execution with --resume"]
            SwarmRun --> BatchResume
        end

        DirectExec["Execute localized edits directly"]
    end

    DAGManifest --> GraphExec
    TaskPacket --> DualLoopExec
    SwarmJob --> SwarmExec
    DirectPlan --> DirectExec

    subgraph Phase4["4. Supervised Verification & Closure"]
        direction TB
        DiffAudit["Diff Audit & Verifier<BR>(scripts/agent_orchestrator.py verify)"]
        TestGate["Mechanically Run Automated Tests<BR>(pytest / project test suites)"]
        Retro["Post-Run Retrospective Survey<BR>(post_run_survey.md)"]
        OuterCommit["Outer Orchestrator Commits & Closes Session<BR>(Sole commit authority)"]

        DiffAudit --> TestGate --> Retro --> OuterCommit
    end

    GraphExec --> DiffAudit
    DualLoopExec --> DiffAudit
    SwarmExec --> DiffAudit
    DirectExec --> DiffAudit

    style Phase1 fill:#e8f0fe,stroke:#1a73e8,stroke-width:2px
    style Phase2 fill:#fef7e0,stroke:#f9ab00,stroke-width:2px
    style Phase3 fill:#e6f4ea,stroke:#137333,stroke-width:2px
    style Phase4 fill:#fce8e6,stroke:#c5221f,stroke-width:2px
    style GraphBranch fill:#e8f0fe,stroke:#1a73e8
    style DualLoopBranch fill:#fef7e0,stroke:#f9ab00
    style SwarmBranch fill:#e6f4ea,stroke:#137333
    style DirectBranch fill:#f1f3f4,stroke:#5f6368
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall