ワンクリックで
runtime-cleanup-worker
Simplify `src/fleet_rlm` internals while preserving or intentionally cleaning up validated public contracts.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Simplify `src/fleet_rlm` internals while preserving or intentionally cleaning up validated public contracts.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Delegate recursive work to child RLM sandboxes with budget management. Use when decomposing tasks into sub-queries, fanning out batched work, managing LLM call budgets, or building parent-child RLM hierarchies.
Diagnose fleet-rlm runtime failures, API contract drift, sandbox errors, and observability issues. Use when something is broken — Daytona connection failures, websocket mismatches, escalation not triggering, budget exhaustion, or missing traces.
Optimize fleet-rlm DSPy programs and RLM skill bundles using GEPA with MLflow tracking. Use when running optimization loops, designing feedback metrics, building training datasets, or comparing runs.
Hub skill for fleet-rlm: when to use dspy.RLM vs ReAct/CodeAct and which workflow skill to load next.
Design DSPy signatures and compose runtime modules for fleet-rlm tasks. Use when creating input/output field definitions, choosing between built-in signatures, selecting execution modes, or wiring custom modules.
Fetch and inspect JavaScript-heavy pages with Playwright in a Daytona browser-capable snapshot.
| name | runtime-cleanup-worker |
| description | Simplify `src/fleet_rlm` internals while preserving or intentionally cleaning up validated public contracts. |
NOTE: Startup and cleanup are handled by worker-base. This skill defines the work procedure for simplification features.
Use this skill for features that:
api, runtime, integrations, and cliagent-browser — invoke when the assigned feature touches browser-shell serving, app navigation, or any behavior validated through the API-served shell.mission-worker-base already ran .factory/init.sh and the baseline commands.test command from .factory/services.yaml for this session. If either step did not run or failed, run it before feature work and record the result; if it still fails, return to the orchestrator without editing code.mission.md, mission AGENTS.md, .factory/library/architecture.md, .factory/library/environment.md, .factory/library/user-testing.md, and .factory/library/cleanup-audit.md.src/fleet_rlm/api/bootstrap_observability.py unless the assigned feature explicitly requires touching it.openapi.yaml, then run the frontend API sync check.make formatmake lintmake typecheckagent-browser against 127.0.0.1:8100 and verify the relevant surface manually..factory/services.yaml (api_contract, websocket_contract, daytona_contract, cli_contract, or quality_gate).README.md so it reflects the final backend structure and validation path.{
"salientSummary": "Moved backend runtime assembly behind a backend-owned factory and removed a CLI-shaped indirection layer without changing the visible command surface. Focused CLI and API contract checks passed, and the API-served shell still loaded from the local server.",
"whatWasImplemented": "Added characterization coverage for the assembly seam, moved the app/runtime construction path out of CLI-oriented helpers, updated imports to point at the new backend owner module, and removed the obsolete pass-through wrapper. The cleanup preserved `fleet web`, `serve-api`, route registration, and shell serving behavior.",
"whatWasLeftUndone": "",
"verification": {
"commandsRun": [
{
"command": "uv run fleet-rlm --help && uv run fleet-rlm daytona-smoke --help",
"exitCode": 0,
"observation": "Canonical CLI command registration remained intact."
},
{
"command": "uv run pytest -q tests/ui/server/test_api_contract_routes.py tests/ui/server/test_router_runtime.py -m 'not live_llm and not live_daytona and not benchmark'",
"exitCode": 0,
"observation": "API route/runtime contract coverage passed after the cleanup."
},
{
"command": "make format && make lint && make typecheck",
"exitCode": 0,
"observation": "Baseline Python validators stayed green."
}
],
"interactiveChecks": [
{
"action": "Opened the API-served shell on http://127.0.0.1:8100 with agent-browser and captured an annotated screenshot.",
"observed": "The shell still loaded correctly with sidebar/navigation and the composer present."
}
]
},
"tests": {
"added": [
{
"file": "tests/unit/test_ws_runtime_prep.py",
"cases": [
{
"name": "test_build_chat_agent_context_uses_backend_owned_builder",
"verifies": "The backend runtime path no longer depends on a CLI-owned assembly seam."
}
]
}
]
},
"discoveredIssues": [
{
"severity": "medium",
"description": "A remaining package-root compatibility export still has active test reachability and should be cleaned up in a follow-up feature.",
"suggestedFix": "Track the residual export cleanup after the current batch lands and contract tests stay green."
}
]
}