prekuter/dryforge / codex/plugin/skills/ready
ready
VerifiedUnderstand what you mean before anything is built. Takes anything — a one-line idea, notes, a spec, or a mix — reads the project first, asks only what is yours to decide, and writes your intent down for you to approve. Use when the user invokes the `ready` skill. Requires git.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 27.0 KB |
| agents/openai.yaml | config | 233 B |
| references/3-doc-gate.md | markdown | 6.0 KB |
| references/decompose.md | markdown | 11.2 KB |
| references/dependency-calc.md | markdown | 6.1 KB |
| references/elicitation.md | markdown | 18.2 KB |
| references/example-3doc.md | markdown | 7.8 KB |
| references/first-cycle-review.md | markdown | 3.1 KB |
| references/foundation-format.md | markdown | 5.3 KB |
| references/gap-analysis.md | markdown | 4.7 KB |
| references/grounds-gate.md | markdown | 3.6 KB |
| references/intent-completeness.md | markdown | 6.3 KB |
| references/intent-review.md | markdown | 4.0 KB |
| references/output-format.md | markdown | 10.0 KB |
| references/project-design-domain.md | markdown | 4.4 KB |
| references/project-design-technical.md | markdown | 4.2 KB |
| references/project-scoping.md | markdown | 3.5 KB |
| references/review-fidelity.md | markdown | 4.6 KB |
references/intent-review.md
View on GitHub# intent-review.md — the completeness lenses (ELICIT's risk-proportional press)
The lenses ELICIT applies while generating what to understand to press into **what the user didn't say**.
Off-the-shelf plan tools only tidy what the user *said*; the hard problems live in the unsaid. These
lenses are how ELICIT mines it — the **risk-proportional press on the high-stakes slots the
decision-surface accounting enumerates** (`elicitation.md`), not an optional flourish: every
load-bearing slot must pass under them before ELICIT may close. What scales with risk is the *depth/force* per item (a trivial goal gets
a light pass; a high-blast goal gets several diverse lenses pressed hard) — **not** whether the sweep
runs. This is the generative completeness work, done *during* dialogue where the user can still answer;
it is **not** a thing to defer to the gate (deferring it is the reward-hack named in `elicitation.md`).
(The **independent** completeness audits are `intent-completeness.md` — auditing the decision surface
before SPEC — and the `3-doc-gate` on the finished artifact; the session→document fidelity check is
REVIEW(A)'s. This file is the generative lenses ELICIT owns.)
**Floor, not ceiling.** The lenses below are a floor to provoke the press, not a fixed checklist; let
what is actually at risk in *this* intent decide where to push.
## Risk-proportional aim — do not blanket-audit
Pressing the whole intent inflates cost for little gain. Aim where the risk lives:
- **Press the assumptions and the non-derivable decisions** — the things the code/material did not
settle. Code-derived, already-grounded content is low risk — skip it.
- **Confirm the load-bearing technical shape was surfaced.** For greenfield (or any intent the code
did not pin), the persistence approach, interface/delivery form, and any plan-defining technical
choice must actually have been surfaced and settled. An un-surfaced shape is a material gap (only the
user can fill it). (Functional completeness is easy to over-focus on while the whole stack goes
unstated; this is the guard against that.)
- **Probe the output / interface contract for completeness.** Where the intent defines an output, API,
schema, or data model, the *shape downstream consumers (and the executor) depend on* is a recurring
source of late-caught gaps: the entities/fields and their constraints, the exact response/output
keys, the status/enum value sets (and whether two conceptually distinct fields were collapsed into
one), uniqueness/identity rules. A half-pinned contract derails downstream work even when the
behavior reads clear — so check the contract is fully specified, not just the behavior. Whether a
given intent even *has* such a contract is judged at runtime (a pure-CLI tool's output format, a
service's response schema, a library's return type — or nothing). Not a fixed field checklist.
## The three lenses (scale to stakes)
A small, low-blast goal gets one quick pass; a complex, high-blast goal gets these diverse lenses,
pressed harder:
- **Implementer lens** — *what can't I build from this?*
- **User-intent lens** — *what did they likely mean that's unstated?*
- **Edge lens** — *what breaks?*
No fixed number of passes — judgment, not a checklist. These lenses are the *qualitative* companion to
`gap-analysis.md`'s mechanical probes: the probes detect silences from the domain's type; these
lenses press the high-stakes assumptions from a skeptic's angle.
## What a lens raises is still a candidate
Anything a lens surfaces is a **candidate**, not yet a question — it passes through `grounds-gate.md`
(site / why-not-covered / consequence) and is asked with a recommended default attached, exactly like
any other ELICIT candidate. A lens that cannot ground its finding drops it.
## Universality guard
No concrete stack, framework, library, or tool name appears here. The lenses and criteria are
stack-agnostic; what is actually at risk is judged at runtime from the intent and the code.
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 |