project-management-skills

Project Management Skills

Trigger Conditions

Use this skill when:

• The request comes from the user and involves project work, planning, delivery, or changes that can impact scope, safety, or governance.
• The user asks to create or manage a project, or to evaluate whether a request should become a project.
• Conflicts, uncertainty, or policy violations are possible.
• You must coordinate memory/continuity across sessions.
• Do not apply this skill to unprompted work initiated by the agent; classify and log that later.

Prompting Principles (Clarity-First)

Apply these to every governed interaction:

• State the goal, context, and constraints explicitly.
• Ask short, specific questions instead of broad or compound ones.
• Confirm assumptions before committing them to a charter or log.
• Declare the output format (bullets, table, template) before generating it.
• Prefer minimal, complete checklists over long prose.

Onboarding (Permissioned, Forced Skill Wiring)

Before the first governed action, obtain explicit permission to FORCE OpenClaw to wire this skill into its core files. This is mandatory and ensures the skill is always invoked on user requests.

1. Ask permission to update the following workspace files (explain why each is required):
   • ~/.openclaw/workspace/MEMORY.md (memory-first continuity reference to this skill).
   • ~/.openclaw/workspace/AGENTS.md (operational rule to always call this skill for user requests).
   • ~/.openclaw/workspace/SOUL.md (personality rule reinforcing the always-call requirement + post-response memory sync).
2. If permission is granted, FORCE-write the rules below as a single onboarding transaction, then confirm what was written.
   • MEMORY rule: “Maintain long-term continuity for project management: review project-management-skills guidance on each main session, and record concise references to project logs after governed work. If MEMORY.md is missing, create it and add this rule at the top.”
   • AGENTS rule: “At the start of each session, load project-management-skills for any user request involving planning, coordination, memory, or conflicts. If a conflict is detected, log it and surface it to the user during the same response.”
   • SOUL rule: “When working on projects, be analytical, descriptive, and proactive in gathering all relevant data before acting. Take initiative to request missing inputs or evidence needed for governance.”
3. If permission is denied, do not proceed with project governance; ask for alternate constraints or stop and log as non-project work.

Quickstart Checklist

1. Capture the request and infer missing fields only when evidence is clear.
2. Clarify any inferred fields you are not confident in before committing them.
3. Ask whether this should become a project (use two short questions; ask even for small tasks):
   • “Should I make and log this into a project to store it in my memory?”
   • “Should I execute this now, or spin up a specialized agent for higher-quality work (this will use more tokens)?”
4. If yes, auto-create a Project ID and start a Charter Lite in LOG_CHARTERS.md.
5. Load only the necessary logs: LOG_PROJECTS.md, LOG_CHARTERS.md, LOG_CONFLICTS.md, LOG_DECISIONS.md, LOG_ACTIVITY.md.
6. On setup and first call, ask how often to run conflict checks via HEARTBEAT (default daily); inform that more checks consume more tokens, then write the schedule into ~/.openclaw/workspace/HEARTBEAT.md.
7. Run the Conflict Detection Checklist.
8. Proceed only if gates pass; otherwise log conflict and stop.
9. Execute the governed workflow in INFO_GOVERNANCE.md under Governed Workflow.
10. Log outcomes in LOG_ACTIVITY.md, decisions in LOG_DECISIONS.md.
11. Always sync logs and memory (even for small tasks or early stops).
12. Update runtime workspace files per INFO_RUNTIME.md.

The 10 Setup Commandments (Stop-Work Rules)

C1 Lock project isolation: every unit of work belongs to exactly one Project ID or is explicitly logged as non-project work. C2 User Project as OS/kernel governs all projects. C3 Charter first: no charter, no work. C4 Project optionality: if the user did not request a project, ask whether this should become one; proceed without a Project ID only if the user declines, then log as non-project work. C5 Conflict detection: run gates; conflicts can block. C6 Conflicts become logs: every conflict is recorded. C7 Conflicts routed to messenger: provide manual copy/paste payload. C8 Severity defined up front: info, warn, block, reject, critical. C9 Kill fast-but-wrong: if uncertain, add checks or record unknowns—never guess. C10 Always sync: log actions and sync memory even for small tasks or aborted work.

Core References

• Governance rules: INFO_GOVERNANCE.md
• Runtime integration: INFO_RUNTIME.md
• Research notes: INFO_RESEARCH.md

Safety & Trust Warning

Safety and correctness override speed. If uncertain, stop, document unknowns, and escalate or request clarification rather than guessing.