# Send Codex Conductor to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete. ``` ### Upgrade existing ```text I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "codex-conductor", "name": "Codex Conductor", "source": "tencent", "type": "skill", "category": "效率提升", "sourceUrl": "https://clawhub.ai/ShalomObongo/codex-conductor", "canonicalUrl": "https://clawhub.ai/ShalomObongo/codex-conductor", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/codex-conductor", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=codex-conductor", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "SKILL.md", "references/codex-runbook.md", "references/gate-checklists.md", "references/gate-prompts.md", "references/manual-test-templates.md", "references/modes.md" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "slug": "codex-conductor", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-29T12:12:29.090Z", "expiresAt": "2026-05-06T12:12:29.090Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=codex-conductor", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=codex-conductor", "contentDisposition": "attachment; filename=\"codex-conductor-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "codex-conductor" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/codex-conductor" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/codex-conductor", "downloadUrl": "https://openagent3.xyz/downloads/codex-conductor", "agentUrl": "https://openagent3.xyz/skills/codex-conductor/agent", "manifestUrl": "https://openagent3.xyz/skills/codex-conductor/agent.json", "briefUrl": "https://openagent3.xyz/skills/codex-conductor/agent.md" } } ``` ## Documentation ### Codex Orchestrator Coordinate Codex as a disciplined delivery system, not a one-shot generator. ### Core Modes Select both: project_mode greenfield: build from scratch brownfield: onboard and modernize an existing system execution_mode autonomous: proceed automatically when gates pass gated: pause at every gate for user approval ### Governing Principle: Spec-Driven Development No code without a spec. This is non-negotiable. Before any implementation, a written spec must exist with: What is being built Why it's needed Acceptance criteria (testable) Constraints and out-of-scope The coding agent MUST NOT: Guess at requirements Make assumptions about behavior Add unrequested features Invent abstractions not in spec If spec is unclear → STOP and ask. Never guess. See references/spec-driven-development.md for full spec templates and enforcement rules. ### Non-Negotiable Sequence Intake + planning questionnaire Spec creation + approval (specs written BEFORE any code) Docs scaffold + AGENTS.md contract Mode-specific pre-architecture work Architecture + ADR baseline (references specs) Build by vertical slices (each task references spec) Verification against spec acceptance criteria Security/quality gates Release readiness + handover Never skip gates silently. Never implement without a spec. ### Required Resources Read these references before running: references/spec-driven-development.md (MANDATORY FIRST - governs all work) references/planning-questionnaire.md references/modes.md references/gate-checklists.md references/testing-matrix.md references/manual-test-templates.md references/codex-runbook.md references/gate-prompts.md scripts/agent_exec.py references/research-playbook.md (if research_mode=true) ### Scaffolding Initialize project artifacts: python scripts/init_project_docs.py --root --mode This creates/updates: AGENTS.md (project workflow contract) docs/*.md planning/architecture/test/progress/change docs brownfield docs (when mode is brownfield) .orchestrator/status.json (machine-readable state) .orchestrator/context.json (project/execution/research mode context) ### Planning Rules Before anything else, ask the user which coding agent to use (codex | claude | opencode | pi) and fallback agent. Then ask all required questions from references/planning-questionnaire.md. Minimum required answers: mission top user journeys v1 scope hosting target stack preference (or explicit request for recommendation) project_mode execution_mode definition of done acceptance tests If research_mode=true, produce docs/research-notes.md and architecture recommendation before G2. ### Greenfield Must complete before G2: requirements + DoD clarity architecture baseline ADR-0001 with alternatives CI/test baseline plan ### Brownfield Must complete before G2 (and authored by coding agent, not orchestrator): as-is architecture and system inventory dependency map and risk register characterization-test baseline migration strategy + rollback approach compatibility boundaries documented ### Gate Engine Use gates G0 through G7 defined in references/gate-checklists.md. Update gate state via script: python scripts/gate_status.py set --root --gate G3 --state PASS --note "slice-1 verified" Validate status schema: python scripts/gate_status.py validate --root Allowed states: PENDING | IN_PROGRESS | PASS | FAIL | BLOCKED. By default, gate preconditions are enforced (sequence + mode-aware docs checks). ### Validation Rules Use references/testing-matrix.md. Mandatory checks per progression: lint/type/build unit/integration/e2e (as applicable) API contract sanity (if API exists) security baseline docs sync verification Also execute manual test scripts from references/manual-test-templates.md. ### Documentation Rules For each meaningful step: update docs/tasks.md update docs/progress.md append docs/change-log.md update docs/traceability.md record test evidence in docs/test-results.md For user-requested changes, run: python scripts/change_impact.py --root --request "" Then complete all TODOs it emits in impacted docs. ### Codex Execution Pattern Use PTY/background for long runs. Follow command patterns in references/codex-runbook.md. Critical rule: each run executes ONE task, not a whole project in one prompt. For G4, maintain docs/g4-task-plan.md checklist and process tasks one by one. Generate gate-specific prompts with: python scripts/generate_gate_prompt.py --gate --agent --project-mode --execution-mode --research-mode --task "" --spec-ref "" update_docs_step.py is now a fallback utility for recovery/manual bookkeeping only. Primary expectation: the coding agent updates docs directly during each task. Required loop: verify spec exists for the task (no spec = no implementation) launch selected coding agent with spec-driven prompt template coding agent updates docs immediately after task completion (including handoff checklist) coding agent wakes OpenClaw with task summary + where verification steps are documented OpenClaw agent runs verification itself: CLI checks in terminal tools Browser/manual checks in browser tools (for web flows) verify output matches spec acceptance criteria if validations fail, OpenClaw sends exact failures back to coding agent and re-runs fix cycle write final gate status only after validations pass (or mark FAIL/BLOCKED) Enforcement: run_gate.py requires --spec-ref for G3/G4 tasks (implementation gates). run_gate.py requires coding agent + fallback agent context. Each task requires validation evidence (--validate-cmd and/or --ui-review-note). Tasks flagged with --requires-browser-check must include --ui-review-note. status=PASS requires at least one --validate-cmd. status=PASS is blocked when --agent-dry-run is used. For G4, PASS is blocked until docs/g4-task-plan.md has no unchecked tasks. Validation output is recorded in docs/validation-log.md. Coding agent must update docs after every task, including docs/agent-handoff.md. In brownfield mode, G1/G2 fail if onboarding docs are not updated by the coding agent. Coding agent prompts MUST include spec preamble from references/spec-driven-development.md. Any implementation without spec reference = automatic FAIL. In autonomous mode, failed validations trigger automatic fix retries (default: 2) with failure details passed back to coding agent. Optional strict mode: --auto-block-on-retry-exhaust auto-classifies gate as BLOCKED when retries are exhausted. ### Progress Visibility Generate a quick status board: python scripts/progress_dashboard.py --root This summarizes current gate, completion %, blockers, and recent activity. Run a single-task gate step with one command: python scripts/run_gate.py --root --gate G2 --agent codex --fallback-agent claude --project-mode brownfield --execution-mode gated --research-mode true --task "architecture baseline refined for API routing" --status IN_PROGRESS --validate-cmd "npm run -s typecheck" --ui-review-note "N/A for architecture-only task" Mark PASS only after all gate-level checklist items are complete: python scripts/run_gate.py --root --gate G2 --agent codex --task "architecture gate complete" --status PASS --validate-cmd "npm run -s typecheck" For web/UI tasks, require browser verification by OpenClaw agent: python scripts/run_gate.py ... --requires-browser-check --ui-review-note "Verified login + CRUD manually in browser via OpenClaw browser tools" Package distributable skill artifact: python scripts/package_skill.py --skill-dir . --out dist ### End-State Deliverables At completion provide: docs/progress.md at 100% final gate summary from .orchestrator/status.json test result summary + unresolved risks deployment + rollback notes next-iteration backlog If blockers remain, mark as PARTIAL_COMPLETE with explicit blockers and owners. ## Trust - Source: tencent - Verification: Indexed source record - Publisher: ShalomObongo - Version: 1.0.0 ## Source health - Status: healthy - Item download looks usable. - Yavira can redirect you to the upstream package for this item. - Health scope: item - Reason: direct_download_ok - Checked at: 2026-04-29T12:12:29.090Z - Expires at: 2026-05-06T12:12:29.090Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/codex-conductor) - [Send to Agent page](https://openagent3.xyz/skills/codex-conductor/agent) - [JSON manifest](https://openagent3.xyz/skills/codex-conductor/agent.json) - [Markdown brief](https://openagent3.xyz/skills/codex-conductor/agent.md) - [Download page](https://openagent3.xyz/downloads/codex-conductor)