richfrem/agent-plugins-skills / plugins/agent-exploration-cycle/.apm/skills/running-exploration-cycle
running-exploration-cycle
VerifiedRuns the business exploration cycle end to end: starts or resumes a session (intake interview, session brief), enforces the Discovery Plan HARD-GATE, routes the four phases (Discovery Plan, Visual Blueprinting, Prototype, Handoff) to the sibling skills and keeps session state. Use when the user says "let's explore this idea", "I have an idea I want to explore", "help me plan this out", "start discovery", "resume my exploration", "where did we leave off", or starts any conversation about a need, problem or idea before building. Not for hardening an existing prototype (use reengineering-vibe-code) or for engineering build work.
references/hard-gate-enforcement.md
View on GitHub# Technical HARD-GATE Enforcement
This document defines the canonical enforcement rule and redirect text for all exploration-cycle agents.
Reference this file in any agent that can initiate capture, build, or synthesis work.
## Gate Condition
Before any of the following work begins:
- Capturing requirements or business rules
- Building a prototype (any component)
- Running user story derivation
- Dispatching handoff-preparer-agent
- Any Phase 3+ activity
**Check:** Does `exploration/discovery-plans/` contain a plan file that the SME has **approved**, meaning it carries `PLAN_STATUS: APPROVED` or a `## SME Approval` section? (`validate_phase_gate.py 1` checks the same thing.)
- **YES** → read the most recent approved plan as the source of truth. Proceed.
- **NO** (no plan, or only a plan marked `## Plan Status: DRAFT`) → use the exact redirect text below. STOP. Do not offer workarounds.
## Canonical Redirect Text (use verbatim)
> "Before we can build or capture anything I need to understand what we're building first.
> Can we start with a short Discovery Planning Session? It usually takes 10–15 minutes and
> it makes sure everything we create after this is exactly what you need."
## Rules
1. Do NOT offer workarounds ("we can skip the plan", "just give me a rough idea").
2. Do NOT continue below the redirect. Return control to the SME and wait.
3. Do NOT try to infer or reconstruct a Discovery Plan from prior conversation — only a real, SME-approved file in `exploration/discovery-plans/` satisfies the gate. File existence alone is not approval; a DRAFT never satisfies it.
4. If an impatient SME says "just build it" or "skip the planning" — use the redirect text again, unchanged. Politely firm.
---
## Tier 3 Hard Stop — Canonical Redirect
This gate applies when the TierGate assessment in Phase 4 determines **Tier 3 (High Risk)** under the one Tier-3 rule in the handoff guide (`handing-off-to-engineering`): Q3 (high-privilege access) or Q4 (financial/compliance) is "yes", or Q1 and Q2 are both "yes". Q5 (people decisions / bias risk) alone is Tier 2 with ethics review, never Tier 3.
### Canonical Redirect Text (use verbatim)
> "This exploration has assessed as high risk — it needs a formal engineering review before anything is deployed or shared outside this session.
> I'll generate a risk summary now so the right team has what they need to take this forward safely.
> You cannot deploy this directly. The next step is handing the risk summary and handoff package to your engineering or security team."
### Rules
1. Do NOT write the handoff package without generating `tier3-risk-summary.md` first.
2. Do NOT allow the SME to proceed to deployment — the Tier 3 outcome is "Formal engineering cycle required", not "deploy after review".
3. If the SME insists on bypassing: repeat the redirect text. Do not soften it.
4. If the SME does not understand why they're being stopped: explain in plain terms which specific TierGate answer triggered it and what that means.
5. The risk summary is not optional and is not a formality — it is the first section of the handoff package for a Tier 3 outcome.
6. If Q5 is "yes" but Q3 and Q4 are both "no" — this is Tier 2 minimum (ethics review required), NOT Tier 3. Do not invoke the Tier 3 hard stop for Q5-alone cases.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |