melodic-software · kyle-sexton · Aug 26, 2026 · Aug 26, 2026 · Aug 26, 2026 · Aug 26, 2026
diff --git a/docs/CLOUD-SESSIONS.md b/docs/CLOUD-SESSIONS.md
@@ -9,6 +9,12 @@ other pages were last verified 2026-07-30. Per the
 [upstream-drift convention](conventions/upstream-drift/README.md), re-fetch a page before acting
 on it.
 
+## Contents
+
+- [What this is](#what-this-is)
+- [Set up your own (any account, machine, or repo)](#set-up-your-own-any-account-machine-or-repo)
+- [How this repository is set up](#how-this-repository-is-set-up)
+
 ## What this is
 
 - [Claude Code on the web](https://code.claude.com/docs/en/claude-code-on-the-web) runs each

diff --git a/docs/conventions/topic-docs/README.md b/docs/conventions/topic-docs/README.md
@@ -23,8 +23,10 @@ content out of the working directory when it outgrows the task.
 
 This directory is the source of truth: this README (tiers, resolution
 order, slug spec, lifecycle), `topic-docs.schema.json` (the tracked
-concern file's shape), `CHANGELOG.md` (version history), `examples/`
-(one worked slice).
+concern file's shape), `CHANGELOG.md` (version history),
+[examples/worked-slice.md](examples/worked-slice.md) (read for one topic
+slice from birth to graduation, when the lifecycle rules need a concrete
+instance).
 
 ## Why this exists
 

diff --git a/docs/topics/ai-adoption-ladder/design/design-threads.md b/docs/topics/ai-adoption-ladder/design/design-threads.md
@@ -1,5 +1,15 @@
 # Design threads — ai-adoption-ladder
 
+## Contents
+
+- [T1: Autonomous-kickoff trigger layer (#240) — RESOLVED 2026-07-17](#t1-autonomous-kickoff-trigger-layer-240--resolved-2026-07-17)
+- [T2: Runtime sandbox bar for autonomy (#245) — RESOLVED 2026-07-17](#t2-runtime-sandbox-bar-for-autonomy-245--resolved-2026-07-17)
+- [T3: Per-work-class guardrail matrix (#243) — RESOLVED 2026-07-17](#t3-per-work-class-guardrail-matrix-243--resolved-2026-07-17)
+- [T4: Autonomous-runner charter (#244) — RESOLVED 2026-07-17](#t4-autonomous-runner-charter-244--resolved-2026-07-17)
+- [T5: Return accounting (#246) — RESOLVED 2026-07-17](#t5-return-accounting-246--resolved-2026-07-17)
+- [T6: Telemetry unification (#247) — RESOLVED 2026-07-17](#t6-telemetry-unification-247--resolved-2026-07-17)
+- [T7: Standing scheduled routines (#242) — RESOLVED 2026-07-17](#t7-standing-scheduled-routines-242--resolved-2026-07-17)
+
 ## T1: Autonomous-kickoff trigger layer (#240) — RESOLVED 2026-07-17
 
 Decision: the trigger layer is a CONTRACT, not a place — thin per-source signal adapters, each

diff --git a/docs/topics/autonomy-ignition/PLAN.md b/docs/topics/autonomy-ignition/PLAN.md
@@ -4,6 +4,15 @@ Sub-topic of `docs/topics/ladder-climb-roadmap/PLAN.md` Phase I (on `main` since
 Design record: interview rulings (ledger committed in the roadmap topic) +
 `design/design-resolution.md`.
 
+## Contents
+
+- [Brief](#brief)
+- [Plan](#plan)
+- [Blast radius](#blast-radius)
+- [Stress-test summary](#stress-test-summary)
+- [Execution shape](#execution-shape)
+- [Handoff to implementation](#handoff-to-implementation)
+
 ## Brief
 
 Ignite the 2→3 loop: the first standing routine (C2 drain — hourly at ignition, re-bound

diff --git a/docs/topics/context-engineering-claude-5/PLAN.md b/docs/topics/context-engineering-claude-5/PLAN.md
@@ -1,5 +1,42 @@
 # Context engineering for Claude 5 — source absorption and rightsizing pass
 
+## Contents
+
+- [Brief](#brief)
+  - [TLDR](#tldr)
+  - [The design documents](#the-design-documents)
+  - [Goal](#goal)
+  - [Settled](#settled)
+  - [Coordination](#coordination)
+  - [Constraints](#constraints)
+  - [Shape — decided](#shape--decided)
+  - [Open](#open)
+  - [Acceptance criteria](#acceptance-criteria)
+- [Plan](#plan)
+  - [Standards grounding](#standards-grounding)
+  - [Phase 1: Finish the fresh-docs sweep [DONE]](#phase-1-finish-the-fresh-docs-sweep-done)
+  - [Phase 2: Resolve the cross-plugin criteria seam [DONE]](#phase-2-resolve-the-cross-plugin-criteria-seam-done)
+  - [Phase 2.5: Proportionality gate — which detectors survive [DONE]](#phase-25-proportionality-gate--which-detectors-survive-done)
+  - [Phase 3: Specify the criteria edits [SPECIFIED — execution moves to Phase 8]](#phase-3-specify-the-criteria-edits-specified--execution-moves-to-phase-8)
+  - [Phase 4: Define the re-run contract [DONE]](#phase-4-define-the-re-run-contract-done)
+  - [Phase 5: Map each check to its owning plugin [DONE]](#phase-5-map-each-check-to-its-owning-plugin-done)
+  - [Phase 6: Design the detectors and the sweep [PARTIALLY SHIPPED]](#phase-6-design-the-detectors-and-the-sweep-partially-shipped--run-contract--suppression-landed-in-audit-pass-detector-catalog-still-open)
+  - [Gate — re-evaluate before implementation [TODO]](#gate--re-evaluate-before-implementation-todo)
+  - [Phase 7: Bring the setup-skill corpus to its owner doc [AUDITED]](#phase-7-bring-the-setup-skill-corpus-to-its-owner-doc-audited)
+  - [Phase 8: Implement the checks in their owning plugins [PARTIALLY SHIPPED]](#phase-8-implement-the-checks-in-their-owning-plugins-partially-shipped--audit-pass-run-contract-shipped-per-check-detectors-still-open)
+  - [Phase 9: Implement the sweep [PARTIALLY SHIPPED]](#phase-9-implement-the-sweep-partially-shipped--audit-pass-skill--run-contract-shipped-cross-plugin-sweep-wiring-still-open)
+  - [Phase 10: Reconcile, then run against this repository [TODO]](#phase-10-reconcile-then-run-against-this-repository-todo)
+  - [Phase 11: Acceptance gate and PR [TODO]](#phase-11-acceptance-gate-and-pr-todo)
+- [Test strategy](#test-strategy)
+- [Alternatives considered](#alternatives-considered)
+- [Risks and mitigations](#risks-and-mitigations)
+- [Blast radius](#blast-radius)
+- [Open questions](#open-questions)
+- [Handoff to implementation](#handoff-to-implementation)
+  - [User-approval gates](#user-approval-gates)
+  - [Execution shape](#execution-shape)
+  - [Mechanical work](#mechanical-work)
+
 ## Brief
 
 Status: **in progress** — shape decided (a component, not a runbook), seam resolved, proportionality

diff --git a/docs/topics/context-engineering-claude-5/design/checks-and-sweep.md b/docs/topics/context-engineering-claude-5/design/checks-and-sweep.md
@@ -10,6 +10,39 @@ Tasks #19 and #22–#27 (two new checks and four edits to existing ones, per the
 and #28 (the sweep). **The sweep is named `audit-pass`**, invoked as `/claude-config:audit-pass` —
 see "Naming, resolved" below for what that choice paid.
 
+## Contents
+
+- [D1 — cross-surface instruction conflict](#d1--cross-surface-instruction-conflict)
+  - [D1 wears the tier label without the delete-and-watch loop, and that is deliberate](#d1-wears-the-tier-label-without-the-delete-and-watch-loop-and-that-is-deliberate)
+  - [Detect](#detect)
+  - [`AGENTS.md` is affirmatively excluded](#agentsmd-is-affirmatively-excluded)
+  - [D1 reads the expanded surface, not the file on disk](#d1-reads-the-expanded-surface-not-the-file-on-disk)
+  - [Type B ships, and does not ship without the allowlist](#type-b-ships-and-does-not-ship-without-the-allowlist)
+  - [Must NOT flag](#must-not-flag)
+  - [Remediate](#remediate)
+  - [Source](#source)
+  - [An incumbent exists after all, and D1 is scoped around it](#an-incumbent-exists-after-all-and-d1-is-scoped-around-it)
+  - [The inventory D1 depends on, and the native-first gate](#the-inventory-d1-depends-on-and-the-native-first-gate)
+- [The other checks](#the-other-checks)
+- [The sweep](#the-sweep)
+  - [Why it is a component and not a runbook](#why-it-is-a-component-and-not-a-runbook)
+  - [Posture](#posture)
+  - [The exclusion set is derived, never hardcoded](#the-exclusion-set-is-derived-never-hardcoded)
+  - [`/doctor` — prerequisite contract, not a hand-wave](#doctor--prerequisite-contract-not-a-hand-wave)
+  - [Dispatch](#dispatch)
+  - [Coverage disclosure — `OPINION` and `/doctor`](#coverage-disclosure--opinion-and-doctor)
+  - [Verification is designed in, not left to the invoker](#verification-is-designed-in-not-left-to-the-invoker)
+  - [D6 needs a synthetic fixture](#d6-needs-a-synthetic-fixture)
+- [The report and the lanes](#the-report-and-the-lanes)
+- [Threat model — prompt injection against the sweep](#threat-model--prompt-injection-against-the-sweep)
+  - [Sources](#sources)
+  - [T1 — instruction text that targets the auditor, not the model under audit](#t1--instruction-text-that-targets-the-auditor-not-the-model-under-audit)
+  - [T2 — the suppression record as an attack surface](#t2--the-suppression-record-as-an-attack-surface)
+  - [T3 — what the human gate can and cannot catch](#t3--what-the-human-gate-can-and-cannot-catch)
+  - [What this section newly constrains, and who owns it](#what-this-section-newly-constrains-and-who-owns-it)
+  - [What this section does not settle](#what-this-section-does-not-settle)
+- [Open in this phase](#open-in-this-phase)
+
 ## D1 — cross-surface instruction conflict
 
 The deliverable's entire officially-backed payload. It ships as check **I12** in

diff --git a/docs/topics/context-engineering-claude-5/design/official-corroboration.md b/docs/topics/context-engineering-claude-5/design/official-corroboration.md
@@ -6,6 +6,15 @@ reference documentation. Every rule it states is checked here against official d
 Where no doc confirms it, the rule is `OPINION`-tier under the authority axis already defined in
 `claude-config/skills/audit-instructions/reference/criteria.md`.
 
+## Contents
+
+- [The authority axis, written out](#the-authority-axis-written-out)
+- [Official material the source omits, relevant to this work](#official-material-the-source-omits-relevant-to-this-work)
+- [Phase 1 sweep — what the eleven pages settled](#phase-1-sweep--what-the-eleven-pages-settled)
+- [Path scoping — verified empirically, not just read](#path-scoping--verified-empirically-not-just-read)
+- [Pages the plan's list missed](#pages-the-plans-list-missed)
+- [`llms.txt` walk — falsifiability check](#llmstxt-walk--falsifiability-check)
+
 Pages fetched this session:
 
 - <https://code.claude.com/docs/en/commands> — `/doctor`

diff --git a/docs/topics/context-engineering-claude-5/design/proportionality-gate.md b/docs/topics/context-engineering-claude-5/design/proportionality-gate.md
@@ -6,6 +6,26 @@ date: 2026-07-24
 
 # Phase 2.5 — proportionality gate on D1–D7
 
+## Contents
+
+- [The test](#the-test)
+- [Dispositions](#dispositions)
+  - [D5's trigger, split in two](#d5s-trigger-split-in-two)
+- [Escalation](#escalation)
+  - [Re-derived shape](#re-derived-shape)
+  - [Ground-truth verification independently strengthened this](#ground-truth-verification-independently-strengthened-this)
+  - [Reconciling with the Brief's one-anchored-source criterion](#reconciling-with-the-briefs-one-anchored-source-criterion)
+- [Homing map — Phase 5, folded in](#homing-map--phase-5-folded-in)
+  - [The determinism contract, re-verified against check 21](#the-determinism-contract-re-verified-against-check-21)
+- [`OPINION`-tier policy](#opinion-tier-policy)
+- [D1 scope questions, answered](#d1-scope-questions-answered)
+  - [D1's detection rule is narrower than "two instructions differ"](#d1s-detection-rule-is-narrower-than-two-instructions-differ)
+- [Corrections this gate forces on upstream documents](#corrections-this-gate-forces-on-upstream-documents)
+- [The design tier, actually re-derived](#the-design-tier-actually-re-derived)
+- [Deferred tool loading — a remainder that is not D6's](#deferred-tool-loading--a-remainder-that-is-not-d6s)
+- [What this gate does not decide](#what-this-gate-does-not-decide)
+- [Independent verification](#independent-verification)
+
 The record PLAN.md's Phase 2.5 sanity checks assert against. It carries every detector's
 disposition with its reason, the escalation verdict, the operator's decision on that escalation, the
 `OPINION`-tier policy, the D1 scope answers the corroboration document parked here, and the homing

diff --git a/docs/topics/context-engineering-claude-5/design/rerun-contract.md b/docs/topics/context-engineering-claude-5/design/rerun-contract.md
@@ -6,6 +6,27 @@ date: 2026-07-24
 
 # Phase 4 — the re-run contract
 
+## Contents
+
+- [1. Finding identity](#1-finding-identity)
+  - [`surface` — the physical file, never the loading entry point](#surface--the-physical-file-never-the-loading-entry-point)
+  - [`anchor` — content-derived, versioned, and granularity-tagged](#anchor--content-derived-versioned-and-granularity-tagged)
+  - [Excerpt extraction, per surface class](#excerpt-extraction-per-surface-class)
+  - [Registered cluster copies canonicalize before identity](#registered-cluster-copies-canonicalize-before-identity)
+  - [`finding_id`, and why the truncations are what they are](#finding_id-and-why-the-truncations-are-what-they-are)
+  - [Assertions](#assertions)
+  - [Borrowed vocabulary, and two deliberate divergences](#borrowed-vocabulary-and-two-deliberate-divergences)
+  - [Matching across runs — the tiered table](#matching-across-runs--the-tiered-table)
+- [2. Where the report lives](#2-where-the-report-lives)
+- [3. Run state, keying, and concurrency](#3-run-state-keying-and-concurrency)
+- [4. Suppression, per target class](#4-suppression-per-target-class)
+- [5. Mid-run resumability](#5-mid-run-resumability)
+- [6. The idempotence properties](#6-the-idempotence-properties)
+  - [The tiers had to be re-derived first — the two-tier split did not survive contact](#the-tiers-had-to-be-re-derived-first--the-two-tier-split-did-not-survive-contact)
+  - [Inputs to the derived tier that are not the tree](#inputs-to-the-derived-tier-that-are-not-the-tree)
+  - [The properties](#the-properties)
+- [Sanity check](#sanity-check)
+
 Task #35. Idempotence is the Brief's headline acceptance criterion, and after the proportionality
 gate it carries more weight than it did: the argument for the sweep existing at all is that *the
 checks are delegated but the run semantics are not*. If this document lands as prose intent, that

diff --git a/docs/topics/fable-field-guide-audit/dispositions.md b/docs/topics/fable-field-guide-audit/dispositions.md
@@ -6,6 +6,45 @@ under `plugins/playbooks/skills/fable-5/**` was edited; the repo is on `main` wi
 Written in a clean context deliberately. Where this document contradicts a ledger, the contradiction
 is stated in §4 with the evidence, not merged away.
 
+## Contents
+
+- [1. How to read this](#1-how-to-read-this)
+- [2. The four dedup calls](#2-the-four-dedup-calls)
+  - [Judgment 1 — S2 Q4b, S5 C4, S13 F1](#judgment-1--s2-q4b-s5-c4-s13-f1-one-root-cause-and-the-ledgers-put-it-in-the-wrong-cell)
+  - [Judgment 2 — S2 Q2, S5 C8, S7 C6](#judgment-2--s2-q2-s5-c8-s7-c6-s7s-one-missing-mechanism-is-refuted-by-s7s-own-body)
+  - [Judgment 3 — S11 Claim 3 and S13 F4](#judgment-3--s11-claim-3-and-s13-f4-neither-of-s11s-landings-hosts-f4-and-f4-does-not-need-one)
+  - [Judgment 4 — S14 C14.5a](#judgment-4--s14-c145a-independent-and-a-prerequisite-for-the-rest)
+- [3. Three decisions the operator owns](#3-three-decisions-the-operator-owns)
+  - [D1 — The `problem-framing.md:60` trigger](#d1--the-problem-framingmd60-trigger)
+  - [D2 — `context/opus-adaptation.md`](#d2--contextopus-adaptationmd)
+  - [D3 — The contradictions against the divergent-spread tactic (S6 C4)](#d3--the-contradictions-against-the-divergent-spread-tactic-s6-c4)
+  - [D4 — Scope-setting pass as an opener (S6 C5)](#d4--scope-setting-pass-as-an-opener-s6-c5)
+- [4. Corrections to the ledgers](#4-corrections-to-the-ledgers)
+- [5. The remediation set](#5-the-remediation-set)
+  - [Tier 0 — prerequisite](#tier-0--prerequisite)
+  - [Tier 1 — the operator-evaluability cluster](#tier-1--the-operator-evaluability-cluster)
+  - [Tier 2 — the show-moves](#tier-2--the-show-moves)
+  - [Tier 3 — independent, single-home](#tier-3--independent-single-home)
+  - [Tier 4 — low priority](#tier-4--low-priority)
+- [6. The claim ledger](#6-the-claim-ledger)
+  - [S1 — map and territory](#s1--map-and-territory)
+  - [S2 — the four quadrants](#s2--the-four-quadrants)
+  - [S3 — unknown-reduction is the skill](#s3--unknown-reduction-is-the-skill)
+  - [S4 — specificity balance, discovery accelerator, starting point](#s4--specificity-balance-discovery-accelerator-starting-point)
+  - [S5 — blind spot pass](#s5--blind-spot-pass)
+  - [S6 — brainstorms and prototypes](#s6--brainstorms-and-prototypes)
+  - [S7 — interviews](#s7--interviews)
+  - [S8 — references](#s8--references)
+  - [S9 — implementation plans](#s9--implementation-plans)
+  - [S10 — fresh session and implementation notes](#s10--fresh-session-and-implementation-notes)
+  - [S11 — pitches and explainers](#s11--pitches-and-explainers)
+  - [S12 — quizzes](#s12--quizzes)
+  - [S13 — the launch-video worked example](#s13--the-launch-video-worked-example)
+  - [S14 — closing claims](#s14--closing-claims)
+- [7. Audit-originated findings — not article-derived](#7-audit-originated-findings--not-article-derived)
+- [8. Remediations rejected on constraint grounds](#8-remediations-rejected-on-constraint-grounds)
+- [9. Sequencing, if any of this is adopted](#9-sequencing-if-any-of-this-is-adopted)
+
 ## 1. How to read this
 
 **Verdict** (per the brief): `covered` / `partial` / `missing` / `contradicted` /

diff --git a/docs/topics/fresh-eyes-checkpoint-audit/PLAN.md b/docs/topics/fresh-eyes-checkpoint-audit/PLAN.md
@@ -1,5 +1,15 @@
 # fresh-eyes-checkpoint-audit
 
+## Contents
+
+- [Brief](#brief)
+- [Plan](#plan)
+- [Blast radius](#blast-radius)
+- [Stress-test summary](#stress-test-summary)
+- [Execution shape](#execution-shape)
+- [Open questions](#open-questions)
+- [Handoff to implementation](#handoff-to-implementation)
+
 ## Brief
 
 ### TLDR

diff --git a/docs/topics/ladder-climb-roadmap/PLAN.md b/docs/topics/ladder-climb-roadmap/PLAN.md
@@ -1,5 +1,15 @@
 # ladder-climb-roadmap — PLAN
 
+## Contents
+
+- [Brief](#brief)
+- [Plan](#plan)
+- [Blast radius](#blast-radius)
+- [Stress-test summary](#stress-test-summary)
+- [Execution shape](#execution-shape)
+- [Open questions](#open-questions)
+- [Handoff to implementation](#handoff-to-implementation)
+
 ## Brief
 
 Interview complete (25/25 resolved, verification-backed — ledger committed alongside this plan:

diff --git a/docs/topics/loop-engineering-codification/PLAN.md b/docs/topics/loop-engineering-codification/PLAN.md
@@ -1,5 +1,10 @@
 # PLAN — loop-engineering-codification
 
+## Contents
+
+- [Brief](#brief)
+- [Plan](#plan)
+
 ## Brief
 
 ### TLDR

diff --git a/docs/topics/plugin-audit-port/PLAN.md b/docs/topics/plugin-audit-port/PLAN.md
@@ -1,5 +1,16 @@
 # plugin-audit-port
 
+## Contents
+
+- [Brief](#brief)
+- [Plan](#plan)
+- [Blast radius](#blast-radius)
+- [Stress-test summary](#stress-test-summary)
+- [Execution shape](#execution-shape)
+- [Decisions made (gate-passed)](#decisions-made-gate-passed)
+- [Open questions](#open-questions)
+- [Handoff to implementation](#handoff-to-implementation)
+
 ## Brief
 
 ### TLDR

diff --git a/plugins/autonomy/.claude-plugin/plugin.json b/plugins/autonomy/.claude-plugin/plugin.json
@@ -1,7 +1,7 @@
 {
   "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
   "name": "autonomy",
-  "version": "0.22.13",
+  "version": "0.22.14",
   "description": "Governed autonomous agent operation: role-topology, binding-seam, wiring-vs-advisor, telemetry, return-accounting, trigger-dispatch, per-work-class guardrail-matrix, standing-routine-catalog, and design-only runner-charter contracts for climbing the AI-adoption ladder, plus a guided-setup skill that discovers an adopting org's state, writes its schema-versioned binding, wires standards-pinned OTLP emission with a zero-cost file-artifact default, wires human-attested return capture at the task boundary, wires signal adapters with one governed dispatch entrypoint, binds the five-class guardrail matrix to an org's isolation substrates with an in-boundary live-validation probe before recording each fail-closed binding, and stands up standing-routine-catalog classes as scheduled temporal signal adapters behind the one governed queue with free scheduling defaults wired as reviewable changes and each routine's work-class mapping homed on the security surface.",
   "author": {
     "name": "Melodic Software",

diff --git a/plugins/autonomy/CHANGELOG.md b/plugins/autonomy/CHANGELOG.md
@@ -3,6 +3,18 @@
 All notable changes to the `autonomy` plugin are documented here. Format follows
 [Keep a Changelog](https://keepachangelog.com/en/1.1.0/); this plugin uses semantic versioning.
 
+## [0.22.14]
+
+### Changed
+
+- **`setup`'s hub reaches every required leaf directly.** The five template files its context
+  slices depend on (return-capture, isolation-probe, routine-definitions, trigger-adapters,
+  ack-reply) and the guardrails admission-policy/work-classes and runner lifecycle/topology leaves
+  gained direct conditioned pointers; the `reference/guardrails.md` and `reference/routines.md`
+  pointers now name those files as the indexes they are; the bare slice and template citations
+  were standardized to the "when `apply` reaches" conditioned form the other slices already used.
+  Behavior unchanged. Progressive-disclosure audit, deep-nesting and blind-pointer treatments.
+
 ## [0.22.13]
 
 ### Changed