PAD Mode
v1.1.1Turn messy requests into structured plans. PAD Mode (Plan → Act → Deliver) gives your AI agent project management superpowers — automatic task breakdown, liv...
⭐ 0· 151·0 current·0 all-time
by@yipxiyi
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (Plan→Act→Deliver) match the SKILL.md and included plan template. Requested artifacts (plan files, templates, status updates, sub-agents) are appropriate for a planning/execution orchestration skill and no unrelated credentials, binaries, or config paths are requested.
Instruction Scope
Instructions are focused on creating/updating plan documents, getting approvals, and then executing tasks (foreground or by spawning sub-agents). This is coherent, but the SKILL.md assumes platform capabilities (spawning sub-agents, sessions_send, writing 'plans/... .md' files and making tool calls during Act) without specifying low-level undo mechanics (e.g., how to 'undo any actions taken' after skipping approval). Users should note that execution steps may perform file writes, external API calls, or deployments once approved.
Install Mechanism
Instruction-only skill with no install spec or code to download; README suggests optional cloning from GitHub but the skill bundle itself contains only documentation and a template. No remote install URLs or extract operations are present.
Credentials
The skill declares no environment variables, credentials, or config paths. That is proportionate to an instruction-only planning skill. However, actual execution (deployments, third-party integrations) will rely on whatever platform/tool credentials the agent already has access to—those are not requested by this skill.
Persistence & Privilege
always is false and the skill does not request permanent elevated presence. It instructs spawning sub-agents and updating plan files in its workspace, which is consistent with its purpose. Autonomous invocation is allowed by platform default but is not unusual here.
Assessment
This skill appears coherent for structured planning and coordinated execution. Before installing, consider: (1) the skill will create and update plan files in the agent workspace and can spawn sub-agents that may perform actions (file edits, web/API calls, deployments) once you approve — ensure you trust the agent's existing tool/credential access; (2) the SKILL.md instructs 'undo' if approval is skipped but doesn't define how to roll back changes—ask the author how undo is implemented if rollback safety matters; (3) be cautious when approving background execution or deployment tasks that require external credentials (e.g., Railway, cloud providers) — those credentials are not requested by the skill but may be used by the agent if available. If you need tighter control, restrict the agent’s tool/credential access or require human approval for any task that touches external systems.Like a lobster shell, security has layers — review code before you run it.
agent-skillvk9731r2czvrbfk6a2havbv6y5183p6qqlatestvk9778kb7q0bxxnasm0k54fcw9183zcryparallel-executionvk9731r2czvrbfk6a2havbv6y5183p6qqplanvk97axxxx79rkrmtpk1mfy64gc583pv8fplan-modevk97axxxx79rkrmtpk1mfy64gc583pv8fplanningvk9731r2czvrbfk6a2havbv6y5183p6qqproductivityvk9731r2czvrbfk6a2havbv6y5183p6qqproject-managementvk9731r2czvrbfk6a2havbv6y5183p6qqsub-agentsvk9731r2czvrbfk6a2havbv6y5183p6qqtask-managementvk9731r2czvrbfk6a2havbv6y5183p6qqworkflowvk9731r2czvrbfk6a2havbv6y5183p6qq
License
MIT-0
Free to use, modify, and redistribute. No attribution required.