ClawHub

PM — AI Product Manager Skill

v1.0.0

You are the Product Manager. Use this skill whenever someone needs a PM — a business stakeholder submitting a requirement, an engineer waiting for priority d...

⭐ 1· 199·0 current·0 all-time
byLika@lika94
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description (AI Product Manager) match the actual contents: extensive PM guidance and playbooks. The skill is instruction-only, requires no binaries, no env vars, and no installs — all appropriate for a purely advisory PM role.
!
Instruction Scope
Most runtime instructions stay within PM activities (decision-making, writing PRDs, prioritization). However, multiple reference documents instruct the PM to 'scan GitHub PRs', 'check chat channels', and 'run metrics scans' — actions that imply access to external systems. The SKILL.md also explicitly states that the agent cannot directly access live systems (GitHub, Slack, analytics). This is an internal inconsistency: the skill expects external signals but has no mechanism or credentials to fetch them. That may lead the agent to repeatedly ask the user for data or to assume data it doesn't have.
✓
Install Mechanism
No install spec and no code files. Instruction-only skills have minimal install risk; nothing is downloaded or written to disk.
✓
Credentials
The skill requests no environment variables, no credentials, and no config paths. This is proportionate to an advisory PM skill and avoids unnecessary access to secrets or external services.
✓
Persistence & Privilege
always:false and no special persistence or system-wide configuration changes. The skill can be invoked autonomously (default), which is normal; there is no evidence it requests elevated or permanent privileges.
Assessment
This skill is coherent for an AI that acts as a Product Manager: it only contains documents and runtime instructions and doesn't ask for credentials or install code. Two practical things to consider before installing: 1) Data supply: many instructions assume the PM can 'scan GitHub, Slack, analytics, PRs' but the skill also says it cannot access live systems. Decide how you want the agent to get live context (manually paste logs/PR lists/dashboards, or provide connectors). If you expect the agent to act on live systems autonomously, this skill has no built-in mechanism to do that. 2) Autonomy expectations: the skill tells the agent to 'be the PM' and make calls. Autonomous invocation is allowed by default; if you do not want the agent to take initiative without explicit user confirmation, restrict its invocation or adoption policy accordingly. No credentials, installs, or network endpoints are present, so there is no immediate exfiltration risk from the skill itself. If you plan to augment it with connectors (GitHub, Slack, analytics), review those connector permissions separately and limit scope as appropriate.

Like a lobster shell, security has layers — review code before you run it.

B2Bvk97c73ph8sxaqvxw31nn9b2qen839wrkPRDvk97c73ph8sxaqvxw31nn9b2qen839wrkaivk97c73ph8sxaqvxw31nn9b2qen839wrkconsumervk97c73ph8sxaqvxw31nn9b2qen839wrklatestvk97c73ph8sxaqvxw31nn9b2qen839wrkproduct-managementvk97c73ph8sxaqvxw31nn9b2qen839wrkroadmapvk97c73ph8sxaqvxw31nn9b2qen839wrkstartupvk97c73ph8sxaqvxw31nn9b2qen839wrk

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments