# Verification > > - **Type:** Skill - **Install:** `agentstack add skill-aniganti-pm-superpowers-verification` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [aniganti](https://agentstack.voostack.com/s/aniganti) - **Installs:** 0 - **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [aniganti](https://github.com/aniganti) - **Source:** https://github.com/aniganti/pm-superpowers/tree/main/plugins/pm-superpowers/skills/verification ## Install ```sh agentstack add skill-aniganti-pm-superpowers-verification ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Verification Before Completion NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE. This skill exists to prevent thin, placeholder-heavy, or internally contradictory strategy artifacts from being shared with stakeholders. ## When to Use Invoke this skill after any document-producing skill (strategy, competitive-landscape, vrio-analysis, strategic-moat, product-ecosystem, pre-mortem, prioritization, decision-log, stakeholder-alignment, prompt-builder) and before sharing the artifact with stakeholders. ## Process ### Step 1: Load the Artifact Read the document the PM wants to verify. The PM may provide: - A file path to a saved artifact - The content directly in the conversation If no artifact is provided, ask the PM to specify which document they want verified. STOP if no artifact is provided and the PM cannot point to one. There is nothing to verify. ### Step 2: Identify the Artifact Type Determine which skill produced the artifact based on its structure and content: | Artifact Type | Key Indicators | |---|---| | Strategy document | Executive summary, strategic pillars, vision/mission, KPIs | | Competitive landscape | Competitor profiles table, positioning map, whitespace | | VRIO analysis | VRIO matrix table, V/R/I/O assessments | | Strategic moat | 8 moat types with ratings, ethical check | | Product ecosystem | Value chain diagram, integration opportunities, closed loops | | Pre-mortem | Tigers/Paper Tigers/Elephants, action plans | ### Step 3: Run the Verification Checklist Apply the appropriate checklist based on artifact type. #### Universal Checks (Apply to ALL artifacts) 1. **Substance over placeholders** — Are all sections populated with real content? Flag any "TBD", "TODO", "[fill in]", or suspiciously generic text. 2. **Evidence traceability** — Does every recommendation or conclusion trace back to specific evidence cited in the document? Flag any unsupported claims. 3. **Internal consistency** — Are there contradictions between sections? (e.g., problem statement says "enterprise users" but target market says "SMB") 4. **Specificity** — Are user segments, competitors, metrics, and recommendations specific? Flag any vague or generic language that could apply to any product. 5. **Actionability** — Could a reader take concrete next steps based on this document? Or does it end with vague aspirations? 6. **Completeness** — Are all expected sections present? Cross-reference against the skill's output structure. #### Strategy-Specific Checks 7. Problem statement follows WHO/WHAT/WHY format with a named user segment 8. At least 2 strategic pillars with VRIO justification 9. North-star metric is quantifiable with baseline/target 10. KPIs connect directly to strategic pillars 11. FAQ has at least 5 questions with substantive answers #### Competitive Landscape-Specific Checks 7. At least 3 competitors profiled with all columns populated 8. Positioning map has 4+ players 9. At least 3 whitespace opportunities with market size estimates 10. "Do nothing" competitor is included 11. Research sources are dated #### VRIO-Specific Checks 7. At least 5 resources evaluated 8. Every "Yes" has specific supporting evidence 9. No blank cells in the VRIO matrix 10. Recommendations exist for each populated category #### Strategic Moat-Specific Checks 7. All 8 moat types assessed with explicit ratings (None/Emerging/Moderate/Strong) 8. Every rating above "None" has specific evidence (not generic claims) 9. Every moat has at least one deepening opportunity identified 10. Top 3 recommendations are concrete and actionable 11. Overall defensibility verdict is candid and evidence-based 12. Ethical check section is present with genuine reflection #### Product Ecosystem-Specific Checks 7. Value chain is mapped with clear layers (supply, platform, distribution, demand) 8. Integration opportunities (backward and forward) are identified with rationale 9. Portfolio coherence is assessed — products reinforce each other, not just coexist 10. Closed-loop opportunities are identified where applicable 11. Short-term and long-term alignment recommendations are distinct and sequenced #### Pre-Mortem-Specific Checks 7. At least 5 Tigers across urgency levels 8. At least 1 Elephant (teams always have unspoken concerns) 9. Launch-Blocking Tigers have complete action plans (risk, mitigation, owner, date) 10. Risks span at least 4 failure dimensions 11. Analyst-generated risks are present (not just PM-provided ones) ### Step 4: Present the Verification Report Format the report as: ``` ## Verification Report: [Artifact Title] **Artifact Type:** [type] **Verified On:** [date] **Verdict:** PASS / PASS WITH WARNINGS / FAIL ### Issues Found #### Critical (Must Fix Before Sharing) - [Issue]: [Description and location in document] #### Warnings (Recommended Fixes) - [Issue]: [Description and suggestion] #### Notes (Minor Observations) - [Observation] ### Summary [1-2 sentence overall assessment] ``` ### Step 5: Remediation If the verdict is FAIL or PASS WITH WARNINGS: 1. List the specific fixes needed 2. Offer to help the PM address each issue 3. After fixes are made, re-run the verification NEVER issue a PASS verdict if critical issues remain. The PM may choose to share anyway, but the verification must be honest. --- ## Red Flags and Anti-Patterns - **Rubber-stamp verification.** NEVER issue a PASS just because the PM wants to move on. The whole point is honest assessment. - **Verifying your own work uncritically.** If you produced the artifact in a prior skill, approach verification with fresh eyes. Look for the gaps you may have glossed over during creation. - **Skipping the checklist.** NEVER give a subjective "looks good to me." Always run the structured checklist. Checklists catch what judgment misses. ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [aniganti](https://github.com/aniganti) - **Source:** [aniganti/pm-superpowers](https://github.com/aniganti/pm-superpowers) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-aniganti-pm-superpowers-verification - Seller: https://agentstack.voostack.com/s/aniganti - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.