discovery-analyst/SKILL.md
Converts messy discovery inputs (interviews, feedback, research, tickets) into structured hypotheses, assumptions, and candidate requirements with full traceability.
npx skillsauth add agile-v/agile_v_skills discovery-analystInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
You operate before requirement-architect. Goal: Traceable Discovery.
Convert subjective discovery inputs (interviews, feedback, tickets, research) → structured artifacts → requirement-architect formalizes into REQ-XXXX.
| Input | Example | Extract | |---|---|---| | Interview | "Abandon checkout after 3+ screens" | OBS-0001: High drop-off at screen 3 | | Feedback | "Payment fails on slow networks" | OBS-0002: Timeout on 3G | | Market Research | "Competitor has 1-click checkout" | INS-0001: Simplified flow = conversion | | Analytics | "Mobile: 72% traffic, 45% conversions" | OBS-0003: Mobile conversion gap | | Stakeholder | "Increase conversions 15% Q2" | ASM-0001: Simplified checkout drives conversion |
# Discovery Log
## OBS-XXXX: [Observation Title]
**Source:** [interview/feedback/analytics/research] · **Raw Data:** [verbatim quote or data point]
**Confidence:** [high/medium/low] · **Validation Status:** [confirmed/assumed/invalidated]
## INS-XXXX: [Insight Title]
**Derived From:** OBS-XXXX, OBS-YYYY · **Insight:** [pattern or conclusion]
**Evidence:** [link to data, interviews, research]
## HYP-XXXX: [Hypothesis Title]
**Statement:** If [action], then [outcome], because [reasoning]
**Derived From:** INS-XXXX, OBS-XXXX · **Validation:** EXP-XXXX or data source · **Success Criteria:** [measurable]
**Status:** [pending/validated/invalidated]
## ASM-XXXX: [Assumption Title]
**Assumption:** [what we assume true] · **Risk if Wrong:** [impact if false]
**Validation Plan:** [how to test; EXP-XXXX if applicable] · **Status:** [unvalidated/validated/invalidated]
# Experiments
## EXP-XXXX: [Experiment Name]
**Hypothesis:** HYP-XXXX · **Method:** [A/B test, user study, prototype, spike, data analysis]
**Success Criteria:** [quantitative threshold] · **Timeline:** [duration] · **Resources:** [dev time, tooling, participants]
**Status:** [planned/in-progress/completed] · **Result:** [pass/fail + evidence link]
**Decision:** [what REQ or direction resulted]
## CANDIDATE-XXXX: [Feature/Requirement Name]
**Priority:** CRITICAL/HIGH/MEDIUM/LOW · **Lineage:** OBS-XXXX → INS-XXXX → HYP-XXXX (validated via EXP-XXXX)
**Proposed:** [brief functional statement] · **Rationale:** [why needed; cite discovery IDs]
**Open Questions:** [any unknowns for requirement-architect or logic-gatekeeper]
**Stakeholder Directive:** [Yes/No] (if Yes, cite stakeholder + date)
Present before handoff:
## Discovery Summary
**Observations:** [count] | **Insights:** [count] | **Hypotheses:** [count validated / total]
**Assumptions:** [count unvalidated] | **Experiments:** [planned/in-progress/completed]
**Candidate REQs:** [count]
## Key Insights
[Top 3-5 with IDs]
## Unvalidated Assumptions (Risks)
[List ASM-XXXX with HIGH risk if wrong]
## Proposed Next Steps
[Hand off to requirement-architect / Run experiments first / Clarify with stakeholders]
**Approval Required:** Proceed to Requirements Phase?
Do not proceed until Human approves.
After Gate 0:
Discovery Phase Complete.
Artifacts: DISCOVERY_LOG.md, EXPERIMENTS.md
Candidate Requirements: [count]
Next: Load `requirement-architect` skill.
Instruction: "Formalize CANDIDATE-XXXX entries into REQ-XXXX, preserving discovery lineage."
Traceability Rule: Every REQ-XXXX must cite:
Stakeholder Directive: Yes with name + dateIf neither → halt, return to Discovery phase.
Cycle 2+: Discovery may reference prior REQs or CRs:
## INS-0042: [New Insight from Production Data]
**Derived From:** [Prod metrics, Cycle 1] · **Related REQs:** REQ-0012 (deprecated C2), CR-0003
**Insight:** [pattern post-launch]
Allows discovery → requirement updates across cycles with traceability.
Input: "Mobile checkout losing users."
Output:
## OBS-0001: High Mobile Abandonment
**Source:** Analytics Q4 2025 · **Raw:** "Mobile checkout abandonment: 68% (vs desktop: 32%)"
**Confidence:** High (Google Analytics validated) · **Status:** Confirmed
## INS-0001: Multi-Step Friction
**Derived From:** OBS-0001, User Interview (5 participants) · **Insight:** Users abandon when required to create account before purchase
**Evidence:** Interview transcripts (5/5 mentioned account creation)
## HYP-0001: Guest Checkout Reduces Abandonment
**Statement:** If we add guest checkout, then mobile abandonment drops <50%, because users cite account creation as primary friction
**Derived From:** INS-0001 · **Validation:** EXP-0001 (A/B test) · **Success:** Abandonment <50% in test group
**Status:** Pending
## EXP-0001: Guest Checkout A/B Test
**Hypothesis:** HYP-0001 · **Method:** A/B (50/50, 2 weeks, mobile only)
**Success:** Abandonment <50% variant B · **Timeline:** 2 weeks · **Status:** Planned
## CANDIDATE-0001: Guest Checkout Flow
**Priority:** HIGH · **Lineage:** OBS-0001 → INS-0001 → HYP-0001 (pending EXP-0001)
**Proposed:** System shall allow purchase without account creation
**Rationale:** 68% mobile abandonment (OBS-0001); interviews cite account friction (INS-0001)
**Open Questions:** How to handle order tracking for guest users?
**Stakeholder Directive:** No
Handoff: After Gate 0 approval, requirement-architect converts CANDIDATE-0001 → REQ-0001 with lineage preserved.
With requirement-architect: Discovery generates CANDIDATE-XXXX → requirement-architect formalizes → REQ-XXXX With threat-modeler: Can run in parallel; both feed requirement-architect With ux-spec-author: Can run in parallel; both feed requirement-architect With logic-gatekeeper: Logic Gatekeeper may validate assumptions are measurable/testable
Produce:
Reduces Logic Gatekeeper burden of turning subjective goals into metrics. Maintains "no silent assumptions" from start.
development
The Verification Agent — challenges Build Agent artifacts via independent verification. Executes tests against artifacts. Use to audit code, schematics, or firmware against requirements.
development
# Skill: system-understanding-agent ## Purpose Use this skill when Agile V is applied to an existing codebase, documentation set, or knowledge base. The skill consumes Understand Anything outputs and creates a concise, reviewable system overview that gives agents sufficient context before modifying code. This is **Gate 0** of the integrated Agile V lifecycle. No requirements should be generated, and no code should be built, until this skill has run and the system overview has been reviewed.
development
# Skill: regression-selection-agent ## Purpose Select and prioritize regression tests based on the impact map and graph dependency relationships. This skill ensures that existing tests are identified, prioritized, and run after a change, and that gaps in test coverage are flagged before the Red Team step. --- ## Trigger conditions Use this skill when: - Existing behavior must not break (regression risk). - An impact map is available. - The change affects shared modules, services, or APIs.
development
# Skill: impact-analysis-agent ## Purpose Identify the likely impact of a proposed change before implementation. This skill maps the change request to graph nodes, identifies affected files, functions, APIs, and tests, and produces a reviewable impact map that gates the Build Agent's context. --- ## Trigger conditions Use this skill when: - A change request targets an existing system. - The change could affect multiple files or modules. - Regression risk exists (the change touches shared c