devrites 6.0.4 → 6.1.1

package/pack/generated/devin/agents/overhaul-recon.md

@@ -15,6 +15,8 @@ allowed-tools:
 
 Do not invoke another agent. You are called by `/overhaul` and return your result to that orchestrator.
 
+Write your receipt (`schema` `overhaul.receipt/1`) to the packet's `receipt_path`, copying `run_id`, `task_id`, `attempt_id`, `lane` and `snapshot` verbatim from the packet and setting `role` to the unprefixed role name (`verifier`, not `overhaul-verifier`); follow `.devin/skills/overhaul/references/orchestration.md` § Receipts, since admission rejects any mismatch.
+
 The `/overhaul` references cited below live under `.devin/skills/overhaul/references/`
 (`.devin/skills/overhaul/references/` on Codex; the host's own skills directory elsewhere).
 

package/pack/generated/devin/agents/overhaul-specialist.md

@@ -15,6 +15,8 @@ allowed-tools:
 
 Do not invoke another agent. You are called by `/overhaul` and return your result to that orchestrator.
 
+Write your receipt (`schema` `overhaul.receipt/1`) to the packet's `receipt_path`, copying `run_id`, `task_id`, `attempt_id`, `lane` and `snapshot` verbatim from the packet and setting `role` to the unprefixed role name (`verifier`, not `overhaul-verifier`); follow `.devin/skills/overhaul/references/orchestration.md` § Receipts, since admission rejects any mismatch.
+
 The `/overhaul` references cited below live under `.devin/skills/overhaul/references/`
 (`.devin/skills/overhaul/references/` on Codex; the host's own skills directory elsewhere).
 

package/pack/generated/devin/agents/overhaul-verifier.md

@@ -15,6 +15,8 @@ allowed-tools:
 
 Do not invoke another agent. You are called by `/overhaul` and return your result to that orchestrator.
 
+Write your receipt (`schema` `overhaul.receipt/1`) to the packet's `receipt_path`, copying `run_id`, `task_id`, `attempt_id`, `lane` and `snapshot` verbatim from the packet and setting `role` to the unprefixed role name (`verifier`, not `overhaul-verifier`); follow `.devin/skills/overhaul/references/orchestration.md` § Receipts, since admission rejects any mismatch.
+
 The `/overhaul` references cited below live under `.devin/skills/overhaul/references/`
 (`.devin/skills/overhaul/references/` on Codex; the host's own skills directory elsewhere).
 

package/pack/generated/devin/skills/devrites-api-interface/SKILL.md

@@ -5,7 +5,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-lib/reference/standards/core.md"],"triggers":{"afk":["devrites-lib/reference/standards/afk-hitl.md"],"deprecation":["devrites-lib/reference/standards/deprecation.md"],"security":["devrites-lib/reference/standards/security.md","rite-review/reference/security-review.md"]},"workspace":["spec.md","plan.md","tasks.md","state.md","decisions.md"]} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-api-interface` bundles every file named below into one deduplicated read.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-api-interface` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read.
 
 # devrites-api-interface: contract before implementation
 

package/pack/generated/devin/skills/devrites-audit/SKILL.md

@@ -6,7 +6,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-lib/reference/standards/core.md","devrites-lib/reference/parallel-dispatch.md","devrites-lib/reference/standards/agents.md"],"triggers":{"audit-coverage":["devrites-lib/reference/standards/audit-coverage.md"],"architecture-health":["devrites-lib/reference/standards/architecture-health.md"]},"workspace":["spec.md","plan.md","tasks.md","touched-files.md","evidence.md","state.md","decisions.md"],"workspaceByRole":{"performance-reviewer":["spec.md","plan.md","tasks.md","state.md","touched-files.md","evidence.md"],"security-auditor":["spec.md","plan.md","tasks.md","state.md","touched-files.md","evidence.md"],"simplifier-reviewer":["spec.md","plan.md","tasks.md","state.md","touched-files.md"]}} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-audit` bundles every file named below into one deduplicated read.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-audit` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read.
 
 # devrites-audit: read-only audit dispatch
 

package/pack/generated/devin/skills/devrites-browser-proof/SKILL.md

@@ -5,7 +5,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-browser-proof/reference/browser-performance.md","devrites-browser-proof/reference/visual-verdict.md","devrites-frontend-craft/reference/quality-standards.md"],"triggers":{"security":["devrites-lib/reference/standards/security.md"],"devex":["devrites-lib/reference/standards/developer-experience.md"]}} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-browser-proof` bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-browser-proof` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
 
 # devrites-browser-proof: runtime evidence for UI
 

package/pack/generated/devin/skills/devrites-debug-recovery/SKILL.md

@@ -6,7 +6,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-lib/reference/standards/code-navigation.md","devrites-debug-recovery/reference/build-the-loop.md","devrites-debug-recovery/reference/cleanup-and-classify.md","devrites-debug-recovery/reference/hypotheses.md","devrites-debug-recovery/reference/instrumentation.md","devrites-debug-recovery/reference/regression-test.md","devrites-debug-recovery/reference/trace.md","devrites-lib/reference/standards/debug-recovery.md"],"triggers":{"one-shot":["devrites-lib/reference/standards/one-shot-actions.md"],"security":["devrites-lib/reference/standards/security.md"],"workspace":["devrites-lib/reference/standards/workflow-artifacts.md"]}} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-debug-recovery` bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-debug-recovery` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
 
 # devrites-debug-recovery: fix the root cause, not the symptom
 
@@ -33,7 +33,8 @@ has no clear next move. Optional slug selects the active workspace; defaults to
    Do not proceed without a reproduction input: the captured error text for
    repeatable actions, or the retained bounded artifact for consumptive ones.
 3. **Ranked hypotheses (3-5, falsifiable):** generate the list before testing
-   any of them. Each must state a prediction.
+   any of them. Each must state a prediction. With a `.devrites/archive/`, first run
+   `devrites-engine recall --paths <failing files>` for prior dead ends and causes.
    **Completion:** 3-5 distinct hypotheses each state an observable prediction.
    See [hypotheses.md](reference/hypotheses.md).
 4. **Trace when ambiguous:** if the cause is unclear, flaky, causal, or one fix
@@ -87,13 +88,14 @@ has no clear next move. Optional slug selects the active workspace; defaults to
   with that fingerprint. Reclassify only on new causal evidence. On cold resume,
   a retained fingerprint with fewer than three such attempts remains runnable
   even if the previous action wrote a terminal cursor.
-- **A maximum of three no-progress attempts per exact causal fingerprint stops the loop.**
+- **A strategy is spent after three no-progress attempts per exact causal fingerprint.**
   Count an attempt only when its recheck preserves the same decisive failure.
   Record attempt number, exact failure, hypothesis, probe, and failed idea after
   each; closure is progress and a different Critical/Important invariant is a new
   fingerprint. There is no JSONL ledger,
   counter command, or reset-on-green operation. Product/acceptance ambiguity, irreversible risk, or
-  human-only access becomes a human gate; otherwise return reproducible `blocked` with
-  `Next: none — technical recovery exhausted for <causal fingerprint>`, never request
-  attempt four. While budget remains, coupled failure requiring behavior change routes
+  human-only access becomes a human gate; otherwise return the spent strategy and
+  reproduction to the caller, which escalates or parks per the
+  [retry contract](../devrites-lib/reference/standards/afk-hitl.md#retry-cap-no-progress-loops-and-self-resolve);
+  never request attempt four of the same strategy. While budget remains, coupled failure requiring behavior change routes
   `/rite-plan repair` inline; behavior-neutral rerouting uses `unblock` inline.

package/pack/generated/devin/skills/devrites-doubt/SKILL.md

@@ -5,7 +5,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-lib/reference/standards/core.md","devrites-lib/reference/standards/agents.md"],"triggers":{"afk":["devrites-lib/reference/standards/afk-hitl.md"],"security":["devrites-lib/reference/standards/security.md"],"tooling":["devrites-lib/reference/standards/tooling.md"]},"workspace":["spec.md","plan.md","tasks.md","state.md","decisions.md"],"workspaceByRole":{"doubt-reviewer":["spec.md","plan.md","tasks.md","state.md","decisions.md"]}} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-doubt` bundles every file named below into one deduplicated read.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-doubt` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read.
 
 # devrites-doubt: CLAIM → EXTRACT → DOUBT → RECONCILE → STOP
 

package/pack/generated/devin/skills/devrites-frontend-craft/SKILL.md

@@ -5,7 +5,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-frontend-craft/reference/craft.md","devrites-frontend-craft/reference/design-references.md","devrites-frontend-craft/reference/quality-standards.md","devrites-frontend-craft/reference/reuse-first.md","devrites-frontend-craft/reference/shape.md","rite-polish/reference/anti-ai-slop.md"],"triggers":{"fullstack":["devrites-frontend-craft/reference/fullstack.md"],"testing":["devrites-lib/reference/standards/testing.md"]}} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-frontend-craft` bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-frontend-craft` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read. Trigger names map to the conditional rules in the sections that follow.
 
 # devrites-frontend-craft: UI like a senior designer-engineer
 

package/pack/generated/devin/skills/devrites-interview/SKILL.md

@@ -5,7 +5,7 @@ triggers:
   - model
 ---
 <!-- loads: {"always":["devrites-lib/reference/standards/core.md","rite-spec/reference/interview-patterns.md"],"triggers":{"afk":["devrites-lib/reference/standards/afk-hitl.md"]},"workspace":["spec.md","questions.md","decisions.md","assumptions.md","decision-coverage.md","state.md"]} -->
-> Read-set manifest: `devrites-engine context [slug] --skill devrites-interview` bundles every file named below into one deduplicated read.
+> Read-set manifest: `devrites-engine context --skills-root .devin/skills <slug> --skill devrites-interview` (no workspace: `--out <path>`) bundles every file named below into one deduplicated read.
 
 # devrites-interview: extract intent
 
@@ -34,7 +34,7 @@ spec, plan, or code.
   **option set** ([[`afk-hitl.md`](../devrites-lib/reference/standards/afk-hitl.md) → "Option set"](../devrites-lib/reference/standards/afk-hitl.md#option-set-how-every-gap-is-presented)): recommended **first**, labelled
   `(Recommended)`, each with a dimension-tagged rationale (`logic · infra · business ·
   architecture`, + `security`/`UX`/`risk` when in scope), plus the escape hatch. Render via
-  `AskUserQuestion` when the harness has it:
+  `ask_user_question` when the harness has it:
   ```
   1. <recommended> (Recommended) — logic: … · infra: … · business: … · architecture: …
   2. <alternative> — <rationale + the trade-off it accepts>

package/pack/generated/devin/skills/devrites-lib/reference/intent-map.md

@@ -69,15 +69,15 @@ Wrong-skill fire: stop, admit it, switch rites.
 | Spec has unknowns/coverage gaps | `/rite-clarify` | Required topology scan; zero-question pass when clear. |
 | Derive intent from existing code without an active feature contract | `/rite-adopt` | Establish the brownfield contract. |
 | Reconcile implementation with an active feature contract | `/rite-converge` | Identify gaps and add missing slices. |
-| Older workspace cannot resume | `/rite-upgrade` | Audit cited current-contract defects; age alone is no defect. Route semantic repairs to normal owners; preserve history without synthetic proof. |
+| Older workspace cannot resume (explicit only) | `/rite-upgrade` | Audit cited current-contract defects; age alone is no defect. Route semantic repairs to normal owners; preserve history without synthetic proof. |
 | Turn an approved spec into its first plan | `/rite-define` | No `tasks.md` slices yet; writes architecture, slices, traceability. |
 | Reslice, reorder, or repair an existing plan | `/rite-plan` | Slices exist and evidence invalidates them. |
 | Review plan before code | `/rite-vet` | Every plan; depth scales to risk. |
 | Small safe fix | `/rite-quick` | Escalates auth, migration, public API, destructive, ambiguous, or multi-slice work. |
 | Mid-size work, fast lane (explicit only) | `/rite-fast` | Spec + plan + questions once, ≤10 slices, one completeness check, review, polish, proof; risk deepens questions. |
 | Prove UI/runtime | `/rite-prove` + `devrites-browser-proof` | Capture real evidence. |
-| Stuck/unfamiliar | `/rite-zoom-out` | Map before editing. |
-| Teach me | `/rite-explain` | Human learning loop. |
+| Stuck/unfamiliar (explicit only) | `/rite-zoom-out` | Map before editing. |
+| Teach me (explicit only) | `/rite-explain` | Human learning loop. |
 | Decide ship readiness | `/rite-seal` | Decides; never mutates Git. |
 | Execute ship after GO | `/rite-ship` | Requires GO/type-GO. |
 | Unattended lifecycle | `/rite-autocomplete` | Clean baseline/checkpoints; hard gates stop. |

package/pack/generated/devin/skills/devrites-lib/reference/parallel-dispatch.md

@@ -4,7 +4,7 @@ Host dispatch; DevRites governs.
 
 ## Roster
 
-Review: Spec/Code; Seal: all applicable roles.
+Review: Spec/Code plus triggered frontend/security/performance; Seal: all applicable roles.
 
 | Reviewer | Fires | Trigger |
 |---|---|---|
@@ -34,12 +34,13 @@ If the candidate diff has no matching surfaces, record
    unverified independence is a gap. Do not launch a probe agent per task;
    still validate each result's context and candidate.
 2. Build each reviewer's read-set once with
-   `devrites-engine context <slug> --phase <p> --role <role> --trigger <set>` and
+   `devrites-engine context --skills-root .devin/skills <slug> --phase <p> --role <role> --trigger <set>` and
    hand the bundle path as the packet's source list instead of a file roster.
    Build roles add `--slice <SLICE-ID>` (slice-scoped, slice-named bundle).
    Evaluate the skill's trigger conditions and pass each that fires — a role's
-   Trigger column maps to manifest names: `security-auditor` → `security`,
-   `frontend-reviewer` → `ui`/`frontend`, `performance-reviewer` → `performance`.
+   Trigger column maps to the phase manifest's names: `security-auditor` →
+   `security`; at Review, `frontend-reviewer` → `ui` and `performance-reviewer` →
+   `performance`. Seal declares neither; those roles load their own standards.
    The `agents` dispatch contract needs no flag: `--role` auto-fires it
    (`auto=[agents]` in the output). The command prints `unselected=[...]`; an
    omitted applicable trigger is a gap, not a shortcut.
@@ -97,7 +98,7 @@ in `review.md`/`seal.md`.
   the same wave (`dispatch open ... --role code-reviewer.frontend --role
   code-reviewer.backend`; distinct handles — the sealed wave's handles and start
   stamps are the concurrency receipt). Build each bundle with `devrites-engine
-  context <slug> --phase <p> --role code-reviewer --trigger <set> --out
+  context --skills-root .devin/skills <slug> --phase <p> --role code-reviewer --trigger <set> --out
   ctx/<p>-code-reviewer.<lane>.bundle.md`; the default name collides. Each packet
   names its lane, the lane's path set (the whole diff stays readable for context),
   the triggered profile ([`review/README.md`](standards/review/README.md)), and the

package/pack/generated/devin/skills/devrites-lib/reference/reply-contract.md

@@ -10,7 +10,7 @@ When an active rite is the controlling caller, Intermediate `NEEDS_REPLAN` (the
 token for the workspace's `NEEDS REPLAN` readiness value), a nested phase `STOP`, and a
 routine Plan/Vet `Next step` are not eligible completion states. Autocomplete may
 use the shapes below only after its requested rest point or a shared genuine
-human/safety/access/exhausted-recovery stop is reached.
+human/safety/access/parked-recovery stop is reached.
 
 ```text
 Done: <result in one sentence>
@@ -104,23 +104,20 @@ Record: .devrites/archive/<slug>/ship.md
 Claims such as proved, reviewed, sealed, shipped, or complete must point to real
 output or an artifact. A NOT-RUN verification keeps its marker in `Evidence:` or `Open:`
 per [`core.md`](standards/core.md#gate-contract) § Gate contract.
-Use exactly one recommended next action except for
-terminal agent-owned technical exhaustion, which has no runnable action.
+Use exactly one recommended next action.
 
-Use that terminal case only per [`one-shot-actions.md`](standards/one-shot-actions.md):
-three recorded no-progress corrections of the exact fingerprint, or required evidence
-irretrievably absent with **no safe in-scope diagnostic-amplification seam**. A spent
-consumptive-action authorization plus a retained new fingerprint is not terminal. For a
-true terminal case use:
+Agent-owned recovery never ends with `Next: none`. Per the
+[retry contract](standards/afk-hitl.md#retry-cap-no-progress-loops-and-self-resolve),
+a spent strategy escalates inline and a fully escalated fingerprint parks while
+independent work continues; with **no safe in-scope diagnostic-amplification seam**,
+escalate or park instead of stopping. A spent consumptive-action authorization plus a
+retained new fingerprint keeps recovering. Once nothing runnable remains, use:
 
 ```text
-Stopped: Technical recovery exhausted
+Stopped: Technical recovery parked
 Blocking: <causal fingerprint and invariant>
-Attempts: <three failed approaches and decisive reproduction>
-No runnable recovery command: unchanged reinvocation remains blocked
-Next: none — requires new evidence or changed failure conditions
+Attempts: <each strategy, its class, and decisive reproduction>
+Completed: <independent work finished after parking>
+Next: <invoking command> — reopens parked fingerprints with fresh escalation
 Record: <artifact path>
 ```
-
-The terminal cursor marker is `Next: none` exactly as shown; a stored `Next step: none`
-names the same exhausted state.

package/pack/generated/devin/skills/devrites-lib/reference/standards/afk-hitl.md

@@ -28,7 +28,7 @@ The contract uses one sentinel, one queue, and one resume verb.
 ## Run modes
 
 - **HITL (default):** human is present. At a gap/checkpoint the skill **asks inline** via
-  the harness `AskUserQuestion` tool: a ranked **option set** (recommended first, each with
+  the harness `ask_user_question` tool: a ranked **option set** (recommended first, each with
   dimension-tagged rationale; see [Option set](#option-set-how-every-gap-is-presented)). The
   human picks; the skill records the pick to `questions.md` (`answered`) + `decisions.md` and
   **continues in place: no `/rite-resolve` round-trip**. `/rite-resolve` is only for answering
@@ -136,9 +136,10 @@ every completed round (serial slice or parallel integrate)
 
 ## Unattended resource envelope
 
-AFK writer admission needs a bounded input queue, effective slice cap, and valid
-`max_agents`, `max_minutes`, and `max_review_queue`. Existing sentinels
-that declare writer admission but miss/malform these fail closed; cold resume keeps the
+Outside `/rite-autocomplete`, AFK writer admission needs a bounded input queue,
+effective slice cap, and valid `max_agents`, `max_minutes`, and `max_review_queue`.
+Existing sentinels that declare writer admission but miss/malform these fail closed
+(Autocomplete's sentinel omits them as unlimited; see below); cold resume keeps the
 state-owned slice counter. A leftover `expires_at` is ignored and never rewritten.
 Read-only watchers use equivalent native caps from [`loop-operations.md`](loop-operations.md).
 
@@ -211,7 +212,7 @@ Wherever a gap, checkpoint, or non-trivial decision surfaces (`/rite-spec`, `/ri
   be silently dropped, merged, or preselected to fit the UI.
 - Recommend for project conventions, stack, scale, and domain, not a generic default.
 
-**HITL** renders the set via `AskUserQuestion`; the human's pick resolves the gate **in place**.
+**HITL** renders the set via `ask_user_question`; the human's pick resolves the gate **in place**.
 **AFK** auto-picks option 1 for gates it may auto-handle. Record the chosen option verbatim and
 keep the **rejected options in `questions.md`**.
 
@@ -304,13 +305,13 @@ Three shapes:
 ```
 /rite-resolve <qid> "<answer>"
 /rite-resolve --drop <qid> ["<reason>"]
-/rite-resolve --batch <path-to-yaml>
+/rite-resolve --batch <plain-text-file>
 ```
 
 `/rite-resolve` is the canonical writer for **async** resume: a gate that already paused and
 stopped the session (an AFK blocking/escalating/irreversible queue, or a HITL pause the human
 walked away from), plus `--batch`. In an **interactive HITL** session the skill resolves the
-`AskUserQuestion` pick **in place** (the same `questions.md` `answered` write + `state.md`
+`ask_user_question` pick **in place** (the same `questions.md` `answered` write + `state.md`
 clear), so you don't type `/rite-resolve` for gaps you answer live. Both paths flip
 `status: open → answered` and clear `Awaiting human` through the **same `devrites-engine state resolve` writer**:
 one source of truth, two entry points (live pick vs typed verb). Use the writer;
@@ -349,7 +350,7 @@ access, safety and irreversible-action boundaries independently stop work.
   invariant + defect mechanism + minimal reproduction/decisive failure signal.
   DEC/DRIFT IDs, line numbers, wording and splitting one cause create no new budget.
 - **Cap no-progress retries:** three no-progress attempts per exact causal fingerprint
-  across wright/recovery. Only a correction whose narrow recheck leaves that cause
+  and strategy across wright/recovery. Only a correction whose narrow recheck leaves that cause
   open or reproduces its failure consumes an attempt.
   Closing a prior finding with discriminating evidence is progress: resolve it,
   do not charge. A new Critical or Important finding gets its own budget only for
@@ -383,11 +384,30 @@ access, safety and irreversible-action boundaries independently stop work.
 - **Reconcile terminal cursors.** Retained distinct Critical/Important evidence below
   cap resumes despite stale `Next step: none`; age/spent action authority is not exhaustion.
 <!-- workflow-artifact-adapter: {"module":"devrites-lib/reference/standards/workflow-artifacts.md","entry":"unattended root reaches current admitted Workflow Artifact work","action":"invoke classifier; execute returned route without wright/slice charge","return":"saved lifecycle phase/action; no intermediate reply"} -->
-- **Classify exhaustion:** human-owned contract/risk/access gaps open their gate.
-  Otherwise preserve reproduction/dead ends, set `Status: blocked` and
-  `Next step: none — technical recovery exhausted; requires new evidence or changed failure conditions`.
-  Emit no phase command, retry-permission question or `/rite-resolve`.
-  Reinvocation with unchanged cause stays blocked; it never resets the cap.
+- **Escalate a spent strategy; never stop on it.** Three no-progress attempts spend
+  the fingerprint's current *strategy*, not the work. Record that strategy and its
+  attempts under `## Dead ends`, then the controlling caller escalates inline in the
+  same invocation: re-diagnose from the reproduction and dead ends, take the first
+  class below that still has an untried strategy, run `/rite-plan repair` plus
+  narrow Vet for it, and resume with a fresh three-attempt budget. Classes, in order:
+  (1) **shared root cause**: fix the owner every failing path routes through and
+  enforce it across all callers instead of another local guard; (2) **boundary
+  redesign**: acceptance-preserving reslice or redesign of the failing seam;
+  (3) **diagnostic amplification**: add discriminating probes or fixtures, then
+  correct from that evidence. A recorded strategy, or a rewording of one, never
+  repeats. Escalation never weakens tests, validators or acceptance. Each
+  fingerprint gets at most three escalations per activation; record each in
+  `drift.md` with its class and strategy.
+- **Park, finish the rest, then stop once.** Human-owned contract/risk/access gaps
+  open their gate. Otherwise, with every escalation spent, mark the fingerprint
+  `parked` in `drift.md`, keep its reproduction and dead ends, and continue every
+  step that does not depend on it. Only when nothing runnable remains, set
+  `Status: blocked` and
+  `Next step: <invoking command> — technical recovery parked; reinvoke to reopen with fresh escalation`.
+  Emit no retry-permission question or `/rite-resolve`. A genuinely new
+  user-issued activation appends `reopened: <date>` to each parked fingerprint and
+  grants it three fresh escalations, counted from that line; wakes, compaction and
+  nested re-entry of the same run never reset counts.
 - **Resolve agent-owned questions first** from code/docs/decisions. Human communication
   is for blocked environment, handover, inaccessible critical information or missing
   credentials/permission—not writing, testing or review. Blocking/escalating/

package/pack/generated/devin/skills/devrites-lib/reference/standards/agents.md

@@ -41,7 +41,7 @@ Skills name exact fresh roles, omit native fields;
 hosts spawn/wait/deliver. Root MUST NOT advance/claim completion before admitting required results;
 running/orphaned/unavailable = `gap` — no root/generic substitute.
 Where a feature workspace exists, every fresh dispatch to a named `devrites-*`
-role — wave or solo — builds its read-set with `devrites-engine context <slug>
+role — wave or solo — builds its read-set with `devrites-engine context --skills-root .devin/skills <slug>
 --phase <p> --role <role>` and tracks the launch through `devrites-engine
 dispatch <slug> open|start|seal|return` (a solo dispatch is a wave of one;
 start/return auto-record metrics). Without a workspace the plain host dispatch

package/pack/generated/devin/skills/devrites-lib/reference/standards/ci-cd.md

@@ -46,7 +46,8 @@ Keep incomplete or risky behavior disabled behind a flag so deploy and release r
   gates precede exposure.
 - Feature flags separate exposure only when both states are tested, the off path preserves
   current behavior, and disabling the flag stops the risky effect. A flag cannot undo an
-  already destructive schema or data change.
+  already destructive schema or data change. Each flag names its removal trigger (full
+  rollout or kill); a flag past its trigger is dead code ([`deprecation.md`](deprecation.md)).
 
 ## Secrets
 

package/pack/generated/devin/skills/devrites-lib/reference/standards/code-review.md

@@ -33,7 +33,7 @@ If it does not, do not merge it.
    `devrites-engine check windows` flags the mechanical subset (suppressions,
    skip/focus, stubs, empty catch, deleted tests, stripped assertions) against
    `windows.md` waivers; loosened thresholds remain this sweep's judgment.
-3. **Correctness:** logic, edge cases, error paths, race conditions, wrong assumptions. For branching or boundary changes, run the [`edge-case trace`](edge-case-trace.md): relevant probe classes, fixed-set siblings, real wiring, negative intent, and deletion contracts with an evidence disposition.
+3. **Correctness:** logic, edge cases, error paths, race conditions, wrong assumptions. A join, match, upsert, or dedupe keyed on a mutable or non-unique attribute (email, name, title) instead of a stable ID is a finding ([`data-integrity.md`](data-integrity.md)). For branching or boundary changes, run the [`edge-case trace`](edge-case-trace.md): relevant probe classes, fixed-set siblings, real wiring, negative intent, and deletion contracts with an evidence disposition.
 4. **Readability:** names, function size, control flow, intent obvious without the author.
 5. **Architecture:** right seam, coupling/cohesion, fits existing patterns, no premature
    abstraction. Check how it fits the larger system as well as its local behavior.

package/pack/generated/devin/skills/devrites-lib/reference/standards/core.md

@@ -3,7 +3,7 @@
 > Applies when: always; every workspace rite reads it first.
 
 Each workspace rite reads core first. Its `loads:` manifest `always` set is its required
-read-set (`devrites-engine context <slug> --skill <name>` where a workspace exists, else
+read-set (`devrites-engine context --skills-root .devin/skills <slug> --skill <name>` where a workspace exists, else
 read the listed files); a trigger's files load when its named condition holds. Load any
 other rule file from `README.md` on demand.
 
@@ -87,7 +87,7 @@ Verdict discipline — no vacuous verdicts:
 
 ## Caller-owned technical backtracking
 
-When a rite invokes an earlier rite inline to repair an agent-owned technical gap, the original rite stays the controlling caller: a nested `STOP` is a phase boundary, not user-facing. The caller re-reads `state.md`, follows the return cursor/`next_action`, and resumes unless a human-owned, safety, access, budget, or exhausted-recovery stop is active ([Persistence before stopping](#persistence-before-stopping-handoff-discipline)). Derive `exhausted-recovery` from the fingerprint's recorded no-progress attempts, not from a stale `state.md` label; one consumed authorization doesn't exhaust offline recovery from retained new evidence.
+When a rite invokes an earlier rite inline to repair an agent-owned technical gap, the original rite stays the controlling caller: a nested `STOP` is a phase boundary, not user-facing. The caller re-reads `state.md`, follows the return cursor/`next_action`, and resumes unless a human-owned, safety, access, budget, or parked-recovery stop is active ([Persistence before stopping](#persistence-before-stopping-handoff-discipline)). Derive `parked-recovery` from the fingerprint's recorded no-progress attempts and escalations, not from a stale `state.md` label; one consumed authorization doesn't exhaust offline recovery from retained new evidence.
 
 An intermediate `Next step` is cold-resume metadata. Do not ask the human to
 copy routine `/rite-plan repair`, `/rite-vet`, `/rite-build`, or proof-rerun

package/pack/generated/devin/skills/devrites-lib/reference/standards/data-integrity.md

@@ -13,6 +13,10 @@ declared invariants.
 Name before planning implementation:
 
 - the authoritative owner of each fact and every writer;
+- the stable identity key of each record: joins, matches, upserts, and dedupes key on
+  an immutable unique ID, never on a mutable or non-unique attribute (email, name,
+  title, path); **failing case:** two accounts share an email, or a user changes it,
+  and the match merges or orphans their records;
 - uniqueness, referential, ordering, range, and lifecycle invariants;
 - tenant/subject partition keys and authorization boundary;
 - transaction boundary and externally visible commit point;

package/pack/generated/devin/skills/devrites-lib/reference/standards/developer-experience.md

@@ -77,6 +77,11 @@ touches:
   carries meaning by colour alone; a status only colour conveys is a finding.
 - **Docs accuracy:** examples are copy-pasteable and correct, the signature in the docs matches
   the code, and changed behavior updated its docs in the same change ([`documentation.md`](documentation.md)).
+- **Agent use:** a coding agent is now a common consumer. Can it complete the flow with no
+  human step: non-interactive flags or env for every prompt, machine-readable output and
+  distinct exit codes where scripts consume results, setup and auth that run headless, and
+  docs that give commands rather than "click here" UI steps? A step only a human can do is
+  friction; name it.
 
 ## Measure, don't assert
 

package/pack/generated/devin/skills/devrites-lib/reference/standards/loop-operations.md

@@ -38,6 +38,9 @@ Every unattended loop must name:
 2. **Objective:** one active workspace or one read-only external observation.
 3. **Cycle:** one documented skill invocation; no hidden command chain.
 4. **Evaluator:** the skill's existing readiness, proof, review, or watcher verdict.
+   A cycle may add proof but never loosens or retargets what judges it (a test, gate,
+   threshold, benchmark, or fixture) to pass; an unattended optimizer overfits whatever
+   it can touch. An objective with no verifiable evaluator is not loopable; it stays HITL.
 5. **Budget:** every applicable `.devrites/AFK` resource cap.
 6. **Checkpoint:** durable workspace/evidence update before the turn ends; a
    `devrites-engine check regression <slug> --update` ratchet after each durable

package/pack/generated/devin/skills/devrites-lib/reference/standards/one-shot-actions.md

@@ -73,8 +73,9 @@ Keep two budgets separate:
 Cold resume does not make that fingerprint old or exhausted. Derive its offline
 no-progress count from `drift.md` and `evidence.md`; while the count is below the
 cap, resume recovery even if a prior writer stored `blocked` / `Next step: none`.
-That terminal cursor is valid only for a human/safety gate, an actually exhausted
-fingerprint, or a proven absence of any safe retention/amplification design.
+A blocked cursor is valid only for a human/safety gate or a parked fingerprint
+with no independent work left; a missing safe retention/amplification design
+escalates to the next correction class.
 
 A new real attempt is normally admissible only after the affected plan and fixtures
 are re-vetted, the failure condition is shown changed, and any required fresh
@@ -98,7 +99,7 @@ claim the runtime failure condition changed: the independently proven change is
 evidence-acquisition invariant, and the attempt's acceptance signal is the promised
 unique retained boundary (or action success).
 
-Terminal `Next: none` is valid only when no safe in-scope diagnostic-amplification
-seam exists, the required change is human/risk/scope owned, or the exact evidence-gap
-fingerprint exhausts bounded recovery. A missing old artifact alone is not terminal,
-and an amplification attempt never inherits authorization from the failed action.
+When no safe in-scope diagnostic-amplification seam exists, escalate or park per
+[`afk-hitl.md`](afk-hitl.md#retry-cap-no-progress-loops-and-self-resolve); only a
+human/risk/scope-owned change stops for a human. A missing old artifact alone is not
+terminal, and an amplification attempt never inherits authorization from the failed action.

package/pack/generated/devin/skills/devrites-lib/reference/standards/performance.md

@@ -15,6 +15,8 @@ Measure first. An optimization without a measurement is a guess that adds comple
 ## Optimize responsibly
 - Fix the **measured** bottleneck, then **re-measure** to prove the win (before/after).
   An optimization that doesn't move the number is just added complexity: revert it.
+  A win counts only while behavior proof stays green on an unchanged workload; tuning
+  the benchmark, fixture, or workload to the candidate games the number and is a finding.
   Keep compact failed and neutral experiment notes in the existing evidence:
   hypothesis, change, comparable measurements, and keep/revert decision. Do not
   create a separate experiment framework or ledger.

package/pack/generated/devin/skills/devrites-lib/reference/standards/skill-authoring.md

@@ -220,12 +220,17 @@ human approval — owned by [`security.md`](security.md) § Prompt-injection and
 Behavior-shaping prose is code:
 
 1. Baseline without guidance; if it passes, add none.
-2. Run ≥5 fresh-context reps/variant; inspect every flagged run.
+2. Run ≥5 fresh-context reps/variant; inspect every flagged run, and ask that agent which
+   instruction or missing context led it there before editing guidance. The corpus holds a control
+   case that must always pass, past-failure cases, and boundary cases whose right answer is
+   stop, ask, or refuse.
 3. Divergent interpretations require rewrite, not averaging.
 4. Pin host/model/build, corpus, grader, and candidate digest or commit+path. Report tasks/trials,
    arms, same-build A/A noise before A/B, sanitized per-trial verdicts/metrics, invalid/null results,
    variance, process versus job outcome, and supported/unproved claims. Never capture raw transcripts;
    lost grading signal is `cannot_verify`.
+5. On each new model generation, re-run step 1 as an ablation: remove the guidance and rerun its
+   eval. Guidance the model no longer needs is a `/rite-learn` retirement candidate, not a keeper.
 
 CI validates only corpora/deterministic artifacts—never paid sessions or lexical claims.
 
@@ -234,6 +239,13 @@ CI validates only corpora/deterministic artifacts—never paid sessions or lexic
 Delete model-default no-ops. Prefer positive targets; reserve prohibitions for hard
 guards. Fill omitted decisions or mark a deliberate branch.
 
+- Instructions add no capability: replace "be careful" or "CRITICAL" with a check, tool,
+  or engine command the agent can run.
+- State both sides of a trade-off. A one-sided cost ("avoid escalating; it is expensive")
+  is overfit: the agent withholds, deflects, or skips the stop it needed.
+- A defensive line added for one model's failure names that failure, so ablation can
+  retire it when a later model no longer fails.
+
 ## Contribution preflight
 
 Record catalog search, owner gap, evals, host parity, and public/internal surface. Public

package/pack/generated/devin/skills/devrites-lib/reference/standards/testing.md

@@ -106,6 +106,11 @@ discriminates the required result.
   any divergence against the external spec — never by majority vote between
   reviewers, and never because the fixture is committed. A golden nobody re-derived
   is a rumor.
+- **A rewrite or port uses the old implementation as its oracle.** Run old and new on the
+  same inputs (the existing suite, a published conformance suite, recorded requests,
+  screenshots) and diff the results;
+  parity claimed only from tests written for the new code is unproven. **Failing case:** a
+  port passes its new unit tests while the old suite fails against it.
 
 ### Safe perturbation
 
@@ -174,6 +179,10 @@ test exercises; **D**escriptive **A**nd **M**eaningful **P**hrases beat **D**on'
 here. (Deliberately trades against production reuse-first: a fixture that makes the reader
 scroll away to understand the case costs more than the duplication saved.)
 
+Test data is synthetic. A generator or factory builds each named edge case (a user with
+1,000 tickets, an empty tenant); production records never enter tests, fixtures, or agent
+sandboxes ([`security.md`](security.md)).
+
 ## Test doubles: reach for the real thing first
 Prefer, in order: **real > fake > stub > mock**. Use the real collaborator when it's fast and
 deterministic; a **fake** (an in-memory implementation that honors the contract) when it isn't; a
@@ -209,6 +218,10 @@ Select cases from the accepted spec and applicable standards, not a generic coun
   authorization, extra fields, or side effects; run schema-driven fuzzers only against a
   disposable target under external-write approval. **Failing case:** a shared
   `createdAt: Date` compiles on both sides while the client receives a string.
+- A module or layer boundary the plan relies on ("only `clients/` imports the SDK") is
+  enforced by a structural test in the project's existing tooling (an import or
+  dependency rule), not by prose. Its failure message names the allowed route, so the
+  agent fixes the design instead of dodging the rule.
 - Compatibility/delivery work drives both feature-flag states and old/new caller or
   schema combinations. Migration-before-code and code-before-migration order each need a
   declared expected result.