Rite Converge

$rite-converge: compare live code with intent

Read spec.md, plan.md, and tasks.md as the sole source of intent (with .devrites/principles.md as governing constraints), assess what the live codebase implements, and append every unmet piece as a new traceable SLICE-### at the bottom of tasks.md so $rite-build can finish it. Use this for a resumed half-built feature, an adopted codebase that drifted from its derived spec, or a build that stalled mid-slice. Read the active workspace first; if there's no spec.md/plan.md/tasks.md, tell the user which prerequisite skill to run.

This is not a diff tool. $rite-converge compares current code with intent. It does not use git history or compare branches. For a change-scoped review use $rite-review; to prove a finished feature use $rite-prove.

Rules consulted (read on demand from .agents/skills/devrites-lib/reference/standards/)

Pull on demand:

• principles.md: the project invariants (.devrites/principles.md); code that violates a MUST principle is the highest-severity gap and produces a remediation slice.
• spec-grammar.md: buildable acceptance criteria vs ## Success metrics (outcome KPIs the code can't make true and this pass never enqueues); structured ### Requirement: / #### Scenario: blocks, each scenario one behavior to check as built / partial / absent.
• tooling.md: use the primary available code index; cross-check only a named unresolved predicate, then fall back to LSP/file search. Read live code, not artifact claims.
• testing.md: a criterion with code but no covering test is partial, not done.
• repository-topology.md, data-integrity.md, integration-reliability.md: only for triggered applicability rows; missing failure/recovery behavior is partial or absent.

Operating rules

• APPEND-ONLY, never rewrite. The only write to tasks.md is appending new SLICE-### entries. Never rewrite, renumber, reorder, or delete an existing slice (including slices a prior convergence appended). Never edit spec.md or plan.md. Never touch application code: completing the appended slices is $rite-build's job, not this skill's.
• Clean means byte-for-byte unchanged. When the code already satisfies everything, leave tasks.md untouched (no empty convergence header) and report a clean result. Recommend $rite-prove.
• Use artifacts as intent. The spec, plan, tasks, and principles are the contract. If assessing reveals the spec is wrong (the code is right and the requirement is stale), that's Spec Drift: stop and route it through the Spec Drift Guard (rite-build/reference/spec-drift-guard.md) + a recorded decision; never paper over a spec bug by appending a task that changes correct code to match a wrong requirement.
• Partial is not done. Code that exists but is untested, half-wired, or covers only the happy path is an unmet gap: enqueue the remainder, don't round it up.
• Principles are non-negotiable. A live violation of a declared invariant with no recorded exception is the top-severity gap, walked first. Absent/empty principles file → none declared → skip the check gracefully, never block for its absence.
• Scout observes; root classifies and writes. Use agents.md. The evidence scout returns live-code citations only; the controlling chat owns built/partial/absent calls and append-only workspace changes.

Workflow

0. Read .agents/skills/devrites-lib/reference/standards/core.md first (the always-on operating rules), then resolve the active slug, require its state.md, and read the cursor directly.
1. Confirm the gate. Require spec.md + plan.md + tasks.md in the active workspace. If any is missing, STOP and name the prerequisite ($rite-spec for a missing spec, $rite-define for a missing plan/tasks, $rite-adopt to onboard existing code). Do not produce partial output. Require decision-coverage.md with Decision coverage: CLEAR; otherwise STOP → $rite-clarify.
2. Load intent: reference/convergence-assessment.md. From spec.md: buildable AC-### / ### Requirement: scenarios (skip ## Success metrics); from plan.md: architecture decisions + named touch-points (files/components the plan says get built); from tasks.md: existing slices + their Satisfies:; from .devrites/principles.md: the invariants. Completion: every buildable criterion, touch-point, slice output, and principle is in the assessment inventory.
3. Reconcile the map, then read the code. Compare spec.md, tasks.md, test-plan.md, and traceability.md directly: every buildable AC/REQ maps to an existing slice, every slice maps back to real acceptance, every slice's recorded proof still agrees with its test-plan.md row, and the mapped prose preserves the requirement's meaning. Record orphaned, invented, duplicate, or contradictory mappings as gaps. Then read the live code (code-intelligence index per tooling.md); artifact mappings never prove implementation.
4. Assess each unit as built / partial / absent against the live code (the rubric is in reference/convergence-assessment.md): every acceptance criterion / scenario, every plan touch-point, and every existing slice's stated Produces. A principle violated in the current code is its own top-severity gap. Dispatch up to three independent inventory partitions to devrites-evidence-scout on the same live-code snapshot, await their dossiers, then reconcile the cited facts in the root context. Completion: every inventory unit is classified once with live-code evidence.
5. Enqueue the remainder as new slices. For each partial or absent unit, append a ## SLICE-### (continue the numbering after the highest existing id), each with a Satisfies: line tracing to the AC/REQ it closes and a Convergence: <iso> marker line. Dependency-order them after the existing slices; a principle-remediation slice sorts first. If every unit is built → append nothing. Completion: every partial/absent unit has one traceable appended slice, or the file is byte-for-byte unchanged.
6. Write append-only + bookkeeping. Append the slice batch to tasks.md (nothing else in that file changes); update traceability.md directly with rows for only the appended slices, preserving every existing row. Appending a slice changes the plan input, so invalidate the prior vet: update state.md to Phase: plan, Next step: $rite-vet, and set an existing eng-review.md field to Implementation readiness: NEEDS REPLAN. When nothing was unmet, leave the plan/vet verdict untouched and set Next step: $rite-prove. Append decisions.md for any material call.
7. STOP. Report units assessed, built / partial / absent counts, slices appended, and any principle violation found; recommend $rite-vet when slices were appended (or $rite-prove if the code already converged).

Completion evidence (fail-closed)

Before reporting "clean" or recommending $rite-prove, confirm:

• Every buildable AC/REQ in the assessment inventory has a built/partial/absent classification with live-code citation
• tasks.md is byte-for-byte unchanged when clean, or append-only when gaps exist
• traceability.md updated only for appended slices
• No narrative "done" without the checklist above

Failing case: all units marked built but one AC lacks a test or runtime citation → report partial, append slice, route $rite-vet.

Appended slice format

Use the complete canonical slice grammar with one added Convergence: field after Satisfies::

<!-- Convergence 2026-07-07: slices below appended by $rite-converge — live code assessed against intent. -->
## SLICE-014 <name of the unmet capability>
Satisfies: AC-007            # the criterion / scenario this closes
Convergence: 2026-07-07      # marks this as a convergence-appended slice, not an original
<all remaining canonical slice fields>

Mid-flight discipline. Do not rewrite an existing slice, edit source, mark a happy-path-only implementation as built, or add work for a spec you suspect is wrong. See reference/anti-patterns.md.