dsh-github-copilot 0.1.0-alpha.13.dsh.0.2.2.rc.3 → 0.1.0-alpha.15.dsh.0.2.2.rc.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (46) hide show
  1. package/AGENTS.md +171 -29
  2. package/CHANGELOG.md +19 -2
  3. package/CONTRIBUTING.md +5 -5
  4. package/README.md +41 -7
  5. package/README.zh.md +36 -5
  6. package/deployment-baseline.json +7 -2
  7. package/docs/account-management-experience.md +2 -2
  8. package/docs/agent-autonomy-evaluation.md +78 -0
  9. package/docs/agent-readiness.md +19 -3
  10. package/docs/auto-high-cost.md +2 -2
  11. package/docs/auto-review-sampling.md +10 -0
  12. package/docs/auto-task-routing.md +3 -3
  13. package/docs/automatic-model-routing.md +3 -3
  14. package/docs/copilot-accounts.md +18 -18
  15. package/docs/copilot-compaction.md +1 -1
  16. package/docs/image-input-compatibility.md +21 -20
  17. package/docs/manual-compaction-recovery.md +3 -3
  18. package/docs/model-compatibility-acceptance.md +12 -5
  19. package/docs/npm-distribution.md +11 -2
  20. package/docs/official-first-016-alpha2.md +3 -2
  21. package/docs/official-first-020-rc1.md +3 -2
  22. package/docs/official-first-020-rc2.md +4 -0
  23. package/docs/official-first-021-alpha4.md +3 -0
  24. package/docs/official-first-021-alpha5.md +3 -0
  25. package/docs/official-first-021-rc3.md +3 -0
  26. package/docs/plugin-diagnostics.md +42 -5
  27. package/docs/session-search-routing.md +9 -6
  28. package/docs/single-route-migration.md +16 -1
  29. package/lib/client.js +77 -7
  30. package/lib/client.js.map +1 -1
  31. package/lib/diagnostics-analysis.js +6 -4
  32. package/lib/{diagnostics-types-DSXrTSCU.js → diagnostics-types-o02ogaiI.js} +9 -1
  33. package/lib/index.js +191 -5
  34. package/lib/manual-compaction-recovery.js +1 -1
  35. package/lib/remote.js +18 -1
  36. package/lib/types/config.d.ts +6 -2
  37. package/lib/types/diagnostics-analysis.d.ts +14 -0
  38. package/lib/types/diagnostics-collector.d.ts +8 -0
  39. package/lib/types/diagnostics-host.d.ts +5 -3
  40. package/lib/types/diagnostics-remote.d.ts +8 -0
  41. package/lib/types/diagnostics-types.d.ts +33 -2
  42. package/lib/types/native-diagnostics.d.ts +5 -0
  43. package/package.json +4 -1
  44. package/scripts/autonomy-evaluation.mjs +313 -0
  45. package/scripts/release-policy.mjs +2 -1
  46. package/tests/fixtures/autonomy-evaluation.json +71 -0
package/AGENTS.md CHANGED
@@ -36,10 +36,61 @@ Use native DSH tools for goals, background jobs and scoped subagents. Use local
36
36
 
37
37
  These are agent instructions and contract-drift checks, not a filesystem sandbox or a guarantee that every session uses a worktree.
38
38
 
39
+ ### Scoped autonomous execution (#491)
40
+
41
+ **Keep progressing within the user's authorization; ask only at a new boundary.**
42
+ Before asking, check the direct user instructions already available for this task.
43
+ Retain the approved target, purpose, side effects, limits and any revocation in
44
+ session-owned task state. Completing one step, receiving a continuation reminder,
45
+ losing a UI window or finding the user unavailable does not by itself erase that
46
+ authorization. A checkpoint records evidence of consent, not a new permission
47
+ grant; respect any tool that requires fresh direct consent or cannot recover it.
48
+
49
+ Proceed with safe inspection, isolated checks and bounded recovery without routine
50
+ confirmation. An explicit "restart and check" approval covers stopping the approved
51
+ app, starting that same target and inspecting it. If fresh discovery subsequently
52
+ proves it is not running before that check completes, allow at most one recovery
53
+ start under the same authorization, unless the user or tool limits consent to one
54
+ invocation. Do not stop an app that is already running for this recovery, start a
55
+ different target, reinstall, change settings/accounts or broaden the operation.
56
+ Unknown app state calls for read-only discovery, not a guessed launch. Repeated
57
+ exit is a concrete failure to investigate and report, not an unlimited restart loop.
58
+
59
+ Tool consent, refusal and interruption controls take precedence. Report
60
+ "the tool reported input interruption", not "you were operating the app";
61
+ the report does not independently establish human activity or intent. Reobserve
62
+ through the same supported tool before any permitted retry. An interrupted
63
+ attempt is not proof of navigation or product failure. Preserve attempt budgets
64
+ across continuation and compaction; do not reset them by renaming the task,
65
+ switching tools to evade refusal, or receiving an automated completion reminder.
66
+
67
+ At a genuine blocker, state the exact missing permission/capability, consequence
68
+ and next action once, then continue any independent authorized work. Do not repeat
69
+ the same question or unchanged waiting message. If no safe work remains, use the
70
+ host's blocked/awaiting-input mechanism when available; otherwise give one truthful
71
+ partial handoff and wait for changed evidence or user input. A completion reminder
72
+ is not new consent, a retry-budget reset or evidence of success. Never mark an
73
+ unverified requirement passed merely to end the loop.
74
+
75
+ | Situation | Decision |
76
+ |---|---|
77
+ | Approved restart/check; exact app is confirmed stopped | Continue the approved start/check, including the one bounded recovery start when consent permits; do not ask again solely because a previous step finished. |
78
+ | App is already running, target differs, or consent was revoked | Observe the running app; obtain new consent only for the new disruptive action, never infer it from prior installation approval. |
79
+ | Tool reports interruption; user says they did not interact | Keep both facts separate, reread UI and follow the tool's permitted recovery; do not blame the user or bypass the tool. |
80
+ | Genuine approval is missing and the user is absent | Keep that action blocked, do other authorized work, and do not send repeated waiting messages. |
81
+ | Install/import checks pass but loaded version is unknown | Preserve partial status; neither a policy check nor disk evidence proves runtime acceptance. |
82
+
83
+ This policy is exposed by `agent.mjs describe/plan` and checked for drift, not
84
+ enforced by a runtime permission engine. It grants no new live-profile access and
85
+ does not bypass tool approvals, plugin-only ownership, required reviews or CI.
86
+
39
87
  ## Product and architecture
40
88
 
41
89
  ### Confirmed product scope (#455, #456)
42
90
 
91
+ Current source target: official DSH `0.2.2-rc.3`. The exact source commit and
92
+ dependency versions are recorded in `deployment-baseline.json` and `package.json`.
93
+
43
94
  The native enhancement request (#460/#461) adds only reviewer-scoped Responses
44
95
  sampling compatibility and an optional canonical-account hosted-search backend.
45
96
  Use public native ownership/configuration proof and prepared dispatch; never
@@ -71,8 +122,8 @@ change other paths or perform this sync automatically.
71
122
  The user-confirmed direction on 2026-11-14 supersedes conflicting legacy feature
72
123
  requirements below: this plugin is a lightweight GitHub Copilot model integration
73
124
  for DSH Desktop, not a second agent product. The qualification target is the exact
74
- official DSH `0.2.2-rc.3` source and published package graph; it is a candidate,
75
- not supported or delivered until the required CI gates pass.
125
+ official DSH `0.2.2-rc.3` source and published package graph. Each new revision
126
+ requires its own CI and release evidence; a previous release does not qualify it.
76
127
 
77
128
  DSH owns application login, AgentLoop, model selection, Auto, parent-model
78
129
  following, subagents, tools, Web, permissions, history, context/compaction, and
@@ -122,7 +173,24 @@ subagent invocation must be verified on `github-copilot`, not inferred from
122
173
  managed-route tests. Record enhancements that cannot safely attach to native
123
174
  APIs rather than patching Core or misrepresenting support.
124
175
 
125
- The last qualified release line is official DSH `0.2.1-rc.3`; its source/artifact CI passed on Windows/Linux in run `40098031196`, while Desktop artifacts/signatures remain unverified. DSH `0.2.2-rc.3` is the current requested candidate and has not passed this repository's required CI. The previous thirteen pins remain historical evidence, not current support or release gates. Do not treat either prerelease as a production Core upgrade recommendation. DSH's built-in `llm-pi-ai` owns the canonical GitHub Copilot provider, OAuth, grant format, token exchange, refresh, and normal model transport.
176
+ For official DSH `0.2.2-rc.3`, plugin `0.1.0-alpha.14.dsh.0.2.2.rc.3`
177
+ at `5d82df760683eeae7bb0413c946e4d25daef2049` passed the Windows/Linux
178
+ source/artifact CI in [run 43691055529](https://github.com/cloga/dsh-github-copilot/actions/runs/43691055529)
179
+ and was published to GitHub Releases and npm. This evidence is limited to that
180
+ revision; later changes require their own CI and publication evidence.
181
+ Desktop signatures, installation, loaded state, OAuth and supplier calls are not
182
+ established. The `desktopCandidateQualified: false` contract
183
+ tracks the unverified Desktop artifact/live-acceptance layer, not a failed CI.
184
+ Earlier pins are historical evidence, not current support gates. A prerelease
185
+ is not a production Core upgrade recommendation. DSH's built-in `llm-pi-ai`
186
+ owns the canonical GitHub Copilot provider, OAuth, grant format, token exchange,
187
+ refresh, and normal model transport.
188
+
189
+ Historical [#483](https://github.com/cloga/dsh-github-copilot/issues/483):
190
+ releases through alpha.12 could reject native schema-default compatibility
191
+ maps, blocking hosted-search proof and reviewer-only temperature omission.
192
+ The alpha.13 release includes a narrow fix; confirm live behavior independently.
193
+ Do not suggest profile edits as a workaround.
126
194
 
127
195
  The following is a historical architecture inventory, not a requirement to keep
128
196
  every surface as a product feature; apply the confirmed scope above when deciding
@@ -164,6 +232,18 @@ evidence stay unknown. Preserve #385 qualification/revocation; clear, pause,
164
232
  expiry and disposal fence later callbacks. Each diagnostic population, including
165
233
  Auto, is independently collapsed by default; opening it never enables collection.
166
234
 
235
+ Native logical observations (#486) require a separate default-off
236
+ `nativeDiagnosticsEnabled` CAS/readback consent. Observe only the public
237
+ `llm/stream` waterfall and the plugin's native hosted-search operation, without
238
+ changing chunks, transport, retries or Usage. Store fixed hourly counts, coarse
239
+ duration buckets and terminal categories, never content, model/account/Session
240
+ IDs or raw errors. Review purpose is not proof of reviewer identity; neither
241
+ logical observation proves HTTP dispatch, supplier receipt, execution or
242
+ billing. Clear/pause/disposal fence late and unconsumed streams. Native and
243
+ account aggregates share the 14-day, 128-live and 2 MiB limits but remain
244
+ separate in presentation and offline analysis. Older strict versions may reject
245
+ new operation codes on rollback; never delete retained evidence to work around it.
246
+
167
247
  ## Session-policy compaction replay consent (#349, #352)
168
248
 
169
249
  `/copilot-compact visible-history` explicitly authorizes one lossy managed Responses
@@ -217,7 +297,8 @@ Keep the native UI and its canonical quota/search attribution. Existing
217
297
  `github-copilot-preview` behavior and Session account overrides remain
218
298
  separately plugin-owned legacy behavior; they are not required for native
219
299
  switching. Disposal removes only plugin registrations. Native model selections
220
- and explicit `web.searchProvider=github-copilot-hosted` still need user-selected
300
+ remain native; plugin-only model selections and explicit
301
+ `web.searchProvider=github-copilot-hosted` need user-selected
221
302
  replacements before uninstall; missing configured Web backends return
222
303
  `WEB_PROVIDER_CONFIGURED_MISSING`, not automatic fallback. Uninstall leaves the
223
304
  last selected grant in the native canonical record. No automatic
@@ -227,7 +308,7 @@ Login environments (#424) use only pinned native pi-ai OAuth for GitHub.com and
227
308
  validated single-label `tenant.ghe.com` Enterprise Cloud hosts. Preserve host
228
309
  binding through identity, duplicate checks, reauthorization and native refresh/
229
310
  discovery/request auth. Plugin identity rejects redirects; native OAuth redirect
230
- control is unavailable in pi-ai 1.1.0 and must not be claimed or emulated with a
311
+ control is unavailable in pi-ai 1.3.0 and must not be claimed or emulated with a
231
312
  global Fetch patch/second flow. GHES and Enterprise quota remain unsupported.
232
313
  See `src/copilot-login-environment.ts` and `docs/copilot-accounts.md`.
233
314
 
@@ -243,9 +324,9 @@ Account display names survive freshness expiry in bounded Host memory, never cre
243
324
 
244
325
  Additional plugin-owned `github-copilot/account-<opaque-id>` records are written directly by official pi-ai OAuth through public account-bound stores. Only an explicit user-confirmed switch may project a selected normalized grant into the canonical record as specified above; preserve the former canonical grant before the first replacement. Never intercept Core credential reads, introduce a separate credential service, patch Core, or implement another wire/catalog. Native `Models.getAuth()` and DSH `modifyRecord()` remain the refresh and serialization owners.
245
326
 
246
- Settings store only the opaque global account selector and explicit `sessionAccounts` rows. Missing/invalid selectors or records fail explicitly, without another-account fallback. Models manages reauthorization/removal and the global default; Credits may select an already authorized account for this Session's subsequent turns, restore global inheritance or run add-only authorization without selecting it (#305/#318). An explicit choice equal to the default remains explicit. Freeze account ownership before Auto assessment and request admission, holding mutation fences until native `turn/end`/disposal; later Session/default changes never redirect running steps, retries or search. Account-local credentials, metadata, quota, proof cancellation and native adapter composition must remain isolated. Keep one public adapter/controller registration, with no second wire/catalog or Core changes. Reauthorization/removal of a pinned record stays blocked; global-default selection may continue independently.
327
+ Settings store only the opaque global account selector and explicit `sessionAccounts` rows. Missing/invalid selectors or records fail explicitly, without another-account fallback. Models manages account authorization and the global default; Credits may select an already authorized managed-route account for this Session's subsequent turns or restore global inheritance (#305/#318). An explicit choice equal to the default remains explicit. Freeze account ownership before request admission, holding mutation fences until native `turn/end`/disposal; later Session/default changes never redirect running steps, retries or search. Account-local credentials, metadata, quota, proof cancellation and native adapter composition must remain isolated. Keep one public adapter/controller registration, with no second wire/catalog or Core changes. Reauthorization/removal of a pinned record stays blocked; global-default selection may continue independently.
247
328
 
248
- Account selection shows the current account and a compact Switch dropdown with no account search, never a flat unbounded account list. Mark the current choice, scroll long lists internally and place Add at the bottom. Saved-account reauthorization/removal remains in a separate management disclosure. One Models Manage disclosure contains account management and cross-account exact-ID model preferences. Continuation settings belong beside account controls. Persist the initial default-on epoch and later global changes with path-level CAS; only unseeded Sessions born after a known epoch inherit it, never retroactively enroll existing/unknown/forked histories. Normal Client continuation uses only the persistent Session policy; no next-turn-only or temporary-recovery duration choices. Explicit-off account changes require persistent-enable/keep-off/cancel choices, then fresh account revision readback before CAS. Exact replay failures offer loss-disclosed persistent Enable/Cancel when off, diagnostics without repeated authorization when on. Missing policy evidence fails explicitly, never a temporary authorization fallback. Retain old next-turn/recovery strict Remote contracts and Host admission for compatibility; failure-evidence expiry does not disable the persistent policy. No automatic send/retry, history writes, private Settings navigation or inferred encrypted account origin. The approved synthetic mock and current navigation limit are documented in [account management](./docs/account-management-experience.md).
329
+ Account selection shows the current account and a compact Switch dropdown with no account search, never a flat unbounded account list. Mark the current choice and scroll long lists internally; Add and saved-account reauthorization/removal remain in Models, not Credits. One Models Manage disclosure contains account management and cross-account exact-ID model preferences. Continuation settings belong beside account controls. Persist the initial default-on epoch and later global changes with path-level CAS; only unseeded Sessions born after a known epoch inherit it, never retroactively enroll existing/unknown/forked histories. Normal Client continuation uses only the persistent Session policy; no next-turn-only or temporary-recovery duration choices. Explicit-off account changes require persistent-enable/keep-off/cancel choices, then fresh account revision readback before CAS. Exact replay failures offer loss-disclosed persistent Enable/Cancel when off, diagnostics without repeated authorization when on. Missing policy evidence fails explicitly, never a temporary authorization fallback. Retain old next-turn/recovery strict Remote contracts and Host admission for compatibility; failure-evidence expiry does not disable the persistent policy. No automatic send/retry, history writes, private Settings navigation or inferred encrypted account origin. The approved synthetic mock and current navigation limit are documented in [account management](./docs/account-management-experience.md).
249
330
 
250
331
  Use hidden volatile `sessionAccounts` with path-level SettingsForms CAS, readback and restart evidence. Session account Remotes use explicit viewed IDs, native Agent lookup and strict codecs, never automatic Client scope. Credits uses that Session's account for identity/quota, rejecting mismatched or late results. Historical Account is an additive assistant-actions disclosure beside native Usage, captured only after a real ordinary native chunk; bounded Host-lifetime evidence becomes unknown after restart/eviction and is never reconstructed from today's settings. No new durable events, history/default/model/Usage writes, automatic account fallback, replay deletion or message retry. See [the requirements and acceptance gates](./docs/copilot-accounts.md).
251
332
 
@@ -260,8 +341,14 @@ Alpha.25 admits native `subagent/descriptor` v3, already v3 in the retained rc.1
260
341
  - `src/index.ts`: authorization bootstrap, dependency-gated Host entry, settings registration, listener, and `ctx.web` provider composition.
261
342
  - `src/authorization-controller.ts`: sign-in/status/sign-out and route mutation.
262
343
  - `src/copilot-grant.ts`, `src/copilot-auth.ts`: strict grant normalization and narrow pi-ai `CredentialStore` adapter; canonical compatibility and explicitly selected account records follow the account-binding contract.
263
- - `src/client.ts`: `settings.models.provider-card` account UI, a `plugins.bundle.config` provider-routing page with Models/footer/section fallback, and independently managed optional Chat integration.
264
- - `src/web-search-routing-config.ts`, `src/web-search-routing-card.ts`: plugin-owned cross-provider live routing settings and the plugin-detail control surface. The user-facing page exposes only primary and fallback providers. Routing is stored directly under the existing `github-copilot` settings namespace at `searchRouting`, mutating both leaves via nested paths `['searchRouting', 'searchProvider']` and `['searchRouting', 'defaultSearchProvider']` in one CAS call; legacy separate routing namespaces are read as fallback only. Revisions coordinate between routing and legacy `searchModel` resets so subsequent edits do not false-conflict. Hosted-search safety settings remain schema-owned rather than a routine tuning form. The Models fallback retains routing only.
344
+ - `src/client.ts`: account and model controls, the native search-model setting,
345
+ and additive optional Chat integration through public slots.
346
+ - `src/native-copilot-access.ts`, `src/native-copilot-search.ts`,
347
+ `src/native-search-model-card.ts`: proof-bound canonical hosted search in
348
+ DSH's existing Web service and its explicit model choice.
349
+ - `src/web-search-routing-config.ts`, `src/web-search-routing-card.ts`: retained
350
+ legacy routing code and compatibility exports; the old routing UI and facade
351
+ are not mounted current behavior. Never reactivate them implicitly.
265
352
  - `tests/fixtures/reasoning-presentation-core.fixture.ts`: native blank-reasoning rendering acceptance and independent Auto/Usage mount evidence. Do not restore the retired plugin renderer/provenance projection or change messages, replay indexes or Usage.
266
353
  - `src/remote.ts`: Typert Remote contribution. Never add credential payloads here.
267
354
  - `src/copilot-usage-host.ts`, `src/copilot-usage-remote.ts`, `src/copilot-usage-card.ts`, `src/copilot-usage-ui.ts`: bounded Host quota snapshots, an independent strict namespace and reversible composer presentation. No real account fetches in tests, private Client store/DOM injection, quota-specific credential owner or general model wire.
@@ -280,11 +367,12 @@ Alpha.25 admits native `subagent/descriptor` v3, already v3 in the retained rc.1
280
367
  - `src/model-protocol.ts`: explicit Core capability detection and conservative legacy fallback.
281
368
  - `src/responses-reasoning.ts`, `src/responses-reasoning-text.ts`: selected-model effort mapping and public summary assembly.
282
369
  - `src/tool-schema-compat.ts`: Copilot-only prompt-assembly filter for unusable escalation arguments; Goal schemas remain native.
283
- - `src/plan.ts`: Copilot-only, fail-closed hosted-search candidate lifecycle.
284
- - `src/auto-assessment-evidence.ts`: bounded request-local semantic classifier milestones copied into strict ephemeral turn evidence; no prompts, replay, credentials or error bodies. Total includes preparation; adapter start and first text are not HTTP dispatch/byte timings. Keep one configured monotonic deadline (30s default, 1000–120000ms) checked at preparation/native request/result boundaries and 128-token compact output; delayed timer settlement is not a real-time guarantee. The user-approved budget trial is not calibrated latency or proof of past timeout causes. Empty classifier setting retains metadata-driven automatic selection; a nonempty exact account ID preserves proof/exclusion/text/headroom guards, never a virtual Auto ID or silent alternate-model fallback. Both native volatile Config leaves affect future assessments and preserve admitted turns. Prefer eligible supplier-advertised reasoning-off classifiers, but request off only when supplier and public prepared-model effort agree; unsupported native controls stay unchanged. No guessed optimal/latency ranking, benchmark requests, adaptive timeout, cooldown or cross-turn cache without reviewed observations.
285
- - `src/probe.ts`: bounded native-search capability proof.
286
- - `src/wire.ts`, `src/wire-anthropic.ts`: inline hosted-search streaming.
287
- - `src/traditional-search.ts`: `github-copilot-hosted` `ctx.web` provider.
370
+ - `src/plan.ts`, `src/probe.ts`, `src/wire.ts`, `src/wire-anthropic.ts`,
371
+ `src/traditional-search.ts`: retained historical hosted-search and routing
372
+ implementation; current registration is owned by `src/native-copilot-search.ts`.
373
+ - `src/auto-assessment-evidence.ts`: retained historical classifier evidence;
374
+ plugin-owned Auto assessment and routing are not mounted. Preserve bounded,
375
+ content-free evidence without reactivating them.
288
376
  - `src/serialize.ts`, `src/sse.ts`, `src/failure.ts`: protocol conversion and bounded error handling.
289
377
  - `src/response-error-body.ts`, `src/request-body-timeout.ts`, `src/request-upload-evidence.ts`: bounded clone-only HTTP error evidence and safe managed upload-timeout guidance. Request-scoped public Node diagnostic events supply numeric local body-write/header/ALPN evidence only after an exact verified `408/user_request_timeout`; absent events, multiple native requests and work limits stay explicit. Always dispose subscriptions when Fetch settles, without changing its arguments, Response/errors, dispatcher, proxy or retry owner. Local body-write completion is not upload duration, kernel ACK, supplier receipt or execution; its absence does not prove incomplete upload. Native failure classification/retry metadata stay unchanged; final JSON bytes are not tokens or a payload limit. A separate explicit opt-in compression setting is default-off and never activated by a 408; never infer overflow, auto-compact, trim history, switch models or add retries from a 408.
290
378
  - `src/responses-request-compression.ts`: asynchronous `node:zlib` gzip at `Z_BEST_SPEED` only for exact account-proven official Copilot HTTPS `/responses` JSON POSTs through the existing public Fetch seam. `github-copilot.responsesRequestCompression` defaults false; enforce the 256 KiB minimum, 32 MiB compression-work bound and both 5%/4 KiB savings thresholds. Skips retain the original request with a fixed reason, and cancellation/account/deadline fences run before dispatch. No custom Fetch, other protocol, explicit `auto`/WebSocket, DNS/proxy/dispatcher changes, context-admission or replay changes, resend after an encoding rejection, or new retry. Gzip is wire encoding, not compaction or proof of supplier receipt.
@@ -298,6 +386,10 @@ Alpha.25 admits native `subagent/descriptor` v3, already v3 in the retained rc.1
298
386
 
299
387
  ## Non-negotiable invariants
300
388
 
389
+ The following two classifier and context-projection paragraphs are historical
390
+ constraints on retained plugin Auto code, not a current runtime feature or
391
+ authorization to reactivate it.
392
+
301
393
  Auxiliary classifier disable proof accepts explicit supplier `off` or the
302
394
  selected protocol's native `none` (Responses/Completions) or `disabled`
303
395
  (Anthropic), scoped to the isolated classifier model only. Missing/empty effort
@@ -339,13 +431,18 @@ evidence is content-free and ephemeral, never a new persistent population.
339
431
  - Copilot OAuth grant writes must rebuild only pi-ai's documented provider fields as a fresh plain JSON object; unrelated extension values never reach DSH credential storage.
340
432
  - Settings changes are path-level. Never replace the whole `llm-pi-ai` section or unrelated provider profiles.
341
433
  - Native sign-out deletes only `llm-pi-ai/github-copilot` and keeps plugin account records, the selector and route settings. A selected plugin account remains saved and must be explicitly applied again after sign-out; a missing selected slot fails explicitly, never automatic canonical fallback.
342
- - Hosted search serves either an eligible initiating `github-copilot` / managed `github-copilot-preview` route, or provider-owned independent search selected by the web-search router. Independent search preserves a nonempty explicit `searchModel` override; otherwise the Host considers at most three eligible Responses models from current account route facts, in deterministic ID order, without a static model list. Metadata alone is not search proof, and the final user query is sent once after candidate proof, never replayed across models. Both paths require an account-available model and supported search protocol; custom inline transport remains legacy-canonical only. Never substitute the global chat default or another Session's identity. The default `probe: true` path requires successful capability proof; `probe: false` is an explicit trust override, not an implicit fallback. Any request containing a file block, including one nested in tool-result content, must bypass the custom wire through `next()` so Core retains file projection ownership.
434
+ - The explicitly selected `github-copilot-hosted` backend uses the canonical native Copilot account and account-proven Responses models also present in the native catalog. Its configured model or deterministic first eligible model receives one fixed capability probe followed by one user query, without fallback or replay. Enforce the 25-second deadline, four-operation limit and account/configuration/disposal fences; metadata alone is not search proof. The old plugin router and custom inline transport are historical, not active alternatives. If that legacy wire is ever deliberately restored, file blocks, including nested tool-result files, must still delegate through `next()` so Core owns file projection.
343
435
  - Misconfiguration and API drift fail loudly with a named missing seam. Do not fall back to process-local secrets or implicit machine state.
344
436
  - Host, Client, and Remote package entries must stay independently buildable and exported.
345
437
  - The package must self-provide authorization when Core omits it, reuse an existing service without duplicate registration, and never activate the integration body before authorization is available.
346
438
 
347
439
  ## Session and migration ownership (V3, alpha.8, #99)
348
440
 
441
+ The following legacy managed-route/search design is historical. The plugin
442
+ router and managed Auto are retired, and removing the canonical native route is
443
+ not the current direction. Do not execute the old migration proposal; use the
444
+ confirmed scope above and the [current README](./README.md).
445
+
349
446
  - One global Host account supplies many models; selected/history-backed Sessions retain independent model context. Auto-mode native search derives from the captured initiating Session's effective request-header/config or explicit `GenerateOptions`, never another Session or future global default C. Fixed/fallback Copilot search uses the explicit provider-owned `searchModel` override when nonempty, otherwise bounded automatic account-owned Responses selection; it never borrows the Chat model, another Session's selection or the global chat default. `Agent.options` is the activation seed, not selected Session-model evidence; `installModelSelection` overrides request/assembly; Core records the effective `Session.requestHeader().config` before tools. Without proven request context, traditional search is unavailable. A newly selected pending model must not reuse the old request header as current prompt guidance. Per-owner plan caches prevent different-model A/B reuse/cancellation; metadata remains shared.
350
447
  - The ordinary Web search card selects only the primary and fallback providers and saves both routing leaves in one CAS mutation in the existing `github-copilot` namespace at `['searchRouting', 'searchProvider']` and `['searchRouting', 'defaultSearchProvider']`. It must not depend on Copilot model discovery, account credentials or a model selection to save routing. Capture traced Settings/routing Remote faces once per UI registration so parent renders do not reset drafts. Preserve a legacy nonempty `searchModel` override; display it read-only and reset it only through an explicit, separate narrow CAS setting `searchModel: ''`, never as a routing-save side effect. Because routing and `searchModel` share the `github-copilot` namespace and revision, successful mutation of either updates the other's expected revision to avoid false CAS conflicts while preserving unsaved drafts. Saving settings does not prove hosted-search capability.
351
448
  - Cold Chat picker and `/model` `listModels()` ensure the shared managed source without Settings first. Actual cold managed search non-forcing ensures metadata before deriving facts; existing 24h maximum TTL, cooldown, credential, allowlist and capability/probe gates remain.
@@ -356,6 +453,9 @@ evidence is content-free and ephemeral, never a new persistent population.
356
453
 
357
454
  ## Live migration readiness (alpha.9, #101)
358
455
 
456
+ Historical readiness evidence only: it is not an authorization or current plan
457
+ to remove the native provider or migrate settings.
458
+
359
459
  - Use no-argument `githubCopilot.migrationStatus()` for fresh live-Agent evidence; generic `session/list` may be stale and plugin inventory alone has no loaded-version evidence. `src/migration-status.ts` reads public leaves synchronously and reports loaded build `plugin.name/version`, `protocolVersion: 1`, `observedAt`, capability flags and completeness flags. Missing/unknown/incomplete is not safe absence.
360
460
  - Capabilities are `agentsList`, `sessionProjections`, `settingsCas`, `providerRegistry`, `defaultSelection`; `complete.sessions/defaultSelection/routes` describes evidence completeness, not migration approval. Effective selection uses pending projection, request-header config, then default only for genuinely empty Sessions with known projection state. Running `activeRequestSelection` is the latest recorded header, not proven in-flight LLM work. Native effective configuration and native/managed registration are separate flags.
361
461
  - This read invokes no auth status/model discovery, credentials or network and mutates no settings/Sessions. No normal UI/global current-model/search card is added. Nine ordinary authorization/model-preference Remotes retain their codec; the tenth migration Remote uses a separate strict `GitHubCopilotMigrationStatus` codec.
@@ -380,9 +480,22 @@ The retained upstream baselines and current qualification target are:
380
480
  - Official tag `dsh-v0.2.0-rc.2`, commit `639ed015397290b3745d163aafe02ffee4aa3f84`; the signed installed Windows Desktop descriptor identifies the rc.2 shared peer graph.
381
481
  - Official tag `dsh-v0.2.1-alpha.4`, commit `86a402200d95fecef4529d969836f1fe13c6cb0d`; historical alpha.16 source/npm qualification only.
382
482
  - Official tag `dsh-v0.2.1-alpha.5`, commit `1517a4dbca0a2789ff1d41ba439600f8547d23fa`; exact source/npm qualification is separate from unverified candidate Desktop artifacts/signatures.
383
- - Official tag `dsh-v0.2.1-rc.3`, commit `1f81c047fc48fa190a683e469daa699857e9afe8`; prerelease target with a public `WebRuntime` Volatile-provider contract change. Source and published artifacts require exact CI qualification; Desktop artifacts/signatures remain unverified.
384
-
385
- The release admission, development dependency and CI targets are exact official `0.2.1-rc.3`; the earlier thirteen pins remain historical evidence only. CI runs unchanged tagged-source and isolated exact published-npm-artifact fixtures on Windows and Linux. Public APIs require pi-ai `1.1.0`, Cordis `4.0.5`, cordis-plugin-loader `1.0.6`, schemastery `3.18.5` and storage-domain-compatible Zod `4.4.3`. The current fixture must await serialized `agent/created`, avoid new synchronous Session-history reads, and preserve MCP/PTC/workflow/Sandbox/Shell/image-offload/Team ownership. Native gateway fixtures bind an explicit public owner Peer and interactive Client carrier; Session-controller fixtures register native resources. Synthetic identities never authorize live access. Neither source nor synthetic tests prove live endpoints, plugin activation, publication or loaded runtime state. See the [rc.3 qualification evidence](./docs/official-first-021-rc3.md); passing source/artifact CI does not authorize Core implementation rewrites or experimental durable records.
483
+ - Official tag `dsh-v0.2.1-rc.3`, commit `1f81c047fc48fa190a683e469daa699857e9afe8`; historical prerelease with a public `WebRuntime` Volatile-provider contract change.
484
+ - Official tag `dsh-v0.2.2-rc.3`, commit `38a6a5c965aa99ab525161efb7cc5cf1fe62fced`; current source and published-package qualification target.
485
+
486
+ Release admission, development dependencies and CI target the exact current
487
+ Core in `package.json`; prior pins are historical. The exact pi-ai,
488
+ Cordis, loader, schemastery and Zod versions also come from package metadata.
489
+ CI runs unchanged tagged-source and isolated published-artifact fixtures on
490
+ Windows and Linux. Fixtures await serialized `agent/created`, avoid new
491
+ synchronous Session-history reads, and preserve MCP/PTC/workflow/Sandbox/Shell/
492
+ image-offload/Team ownership. Native gateway fixtures bind an explicit public
493
+ owner Peer and interactive Client carrier; Session-controller fixtures register
494
+ native resources. Synthetic identities never authorize live access. Source or
495
+ synthetic tests alone do not prove live endpoints, plugin activation,
496
+ publication or loaded runtime state. The [0.2.1-rc.3 review](./docs/official-first-021-rc3.md)
497
+ is historical; passing source/artifact CI never authorizes Core rewrites or
498
+ experimental durable records.
386
499
 
387
500
  These pins document compatibility evidence. They do not authorize creating another controlled Core patch or making one a prerequisite for new plugin fixes.
388
501
 
@@ -400,9 +513,9 @@ These pins document compatibility evidence. They do not authorize creating anoth
400
513
  - Remote results: the nine ordinary authorization/model-preference methods retain the Zod v4 `GitHubCopilotAuthorizationView` strict codec required by rc.2 and accepted by rc.1. The tenth no-argument `migrationStatus()` method has a separate strict `GitHubCopilotMigrationStatus` codec; it does not change the ordinary auth contract.
401
514
  - Single-model saves use the additive `setModelExcluded(modelId, excluded)` Remote with an independent strict `GitHubCopilotModelPreferencesView` result. Preserve the ten existing descriptor identities/codecs. Do not read credentials, discover models or enumerate Sessions on this mutation path; confirm native CAS persistence before returning.
402
515
 
403
- When upgrading DSH or pi-ai, inspect the exact tagged public exports and update the baseline, compatibility guard, tests, and docs together. Apply the **official-first policy**: compare each customization's purpose against exact official source/contracts, classify complete/partial/unverified support, choose retain/migrate/retire, and record the remaining gap plus a concrete retirement trigger. Unverified parity is not evidence of absence. Prefer official behavior only after configuration/data migration, safety and runtime acceptance are reviewed; remove redundant paths and their obsolete tests without losing user-visible acceptance coverage. The current comparison is [official-first alpha.2](./docs/official-first-016-alpha2.md).
516
+ When upgrading DSH or pi-ai, inspect the exact tagged public exports and update the baseline, compatibility guard, tests, and docs together. Apply the **official-first policy**: compare each customization's purpose against exact official source/contracts, classify complete/partial/unverified support, choose retain/migrate/retire, and record the remaining gap plus a concrete retirement trigger. Unverified parity is not evidence of absence. Prefer official behavior only after configuration/data migration, safety and runtime acceptance are reviewed; remove redundant paths and their obsolete tests without losing user-visible acceptance coverage. The [alpha.2 comparison](./docs/official-first-016-alpha2.md) is historical, not the current release qualification.
404
517
 
405
- For alpha.25, source markers, local rc.1-backed focused tests and fifteen scoped exact-source runtime tests passed (alpha.2 contracts 8, Remote 1 and Session-context 6). Full local `pnpm verify` passed: 1373 Vitest tests with 2 expected skips, 176 tooling tests, typechecks, build and package smoke; pack/tarball verification passed. The scoped tests use a supplemental resolver with official TypeScript `6.0.3`, declared `mime-types@3.0.2` and `ws@8.21.0`, and shared Zod `^4.4.3`, without source/dependency patches, not the full official-root-helper closure. Broad frozen dependency installation is blocked by the configured mirror's HTTP 404 for `node-addon-require-builtin@0.1.6`, and candidate CI qualification has not executed. Keep these limits separate from live Desktop, published-artifact, OAuth and model-call compatibility; none is claimed.
518
+ For historical alpha.25, source markers, local rc.1-backed focused tests and fifteen scoped exact-source runtime tests passed (alpha.2 contracts 8, Remote 1 and Session-context 6). Full local `pnpm verify` passed: 1373 Vitest tests with 2 expected skips, 176 tooling tests, typechecks, build and package smoke; pack/tarball verification passed. The scoped tests used a supplemental resolver with official TypeScript `6.0.3`, declared `mime-types@3.0.2` and `ws@8.21.0`, and shared Zod `^4.4.3`, without source/dependency patches, not the full official-root-helper closure. That historical local mirror lacked `node-addon-require-builtin@0.1.6`; the observation is not the current CI status. Keep historical limits separate from live Desktop, published-artifact, OAuth and model-call compatibility.
406
519
 
407
520
  ## Code and documentation conventions
408
521
 
@@ -455,8 +568,9 @@ plugin version. Attempt safe authorized checks on an already-running Desktop;
455
568
  unreliable local npm is not a reason to skip them. Use existing computer-use
456
569
  tools, never shell activation or a repository script as an App controller.
457
570
  Begin with current app discovery, not a historical "already open" claim. If not
458
- running, start only with explicit launch authorization via controlled `start_app`
459
- using the exact approved target; then verify discovery and fresh UI, not merely
571
+ running, start only with explicit launch authorization (including an applicable
572
+ restart/check approval retained under scoped autonomous execution) via controlled
573
+ `start_app` using the exact approved target; then verify discovery and fresh UI, not merely
460
574
  the launch receipt. Do not start an already-running app again. A welcome screen
461
575
  is not automatically an auth blocker: inspect an official safe skip first. DSH's
462
576
  Add API Key -> Set up later path does not submit/store a key or authenticate;
@@ -464,17 +578,22 @@ use it only when actually exposed, within tool permission rules. Never infer
464
578
  Copilot account invalidity from welcome UI or bypass real authentication/warnings.
465
579
  Welcome dismissal may replace the window: rediscover app/window and read its fresh
466
580
  tree before acting, rather than retrying a stale window on an empty immediate diff.
467
- Honor user-input interruption: reread UI before retrying the same permitted action.
581
+ Honor tool-reported input interruption without inferring human intent: reread UI
582
+ before retrying the same permitted action.
468
583
  Before handoff, freshly discover the current app and inspect the target panel or
469
584
  workspace, reporting the observation time only; past success does not establish
470
585
  current running state or guarantee the app remains running later.
471
586
  An action receipt or empty immediate diff is not a navigation failure: obtain
472
587
  a fresh post-action `get_window_state` observation first. Keep one attempt per
473
- supported navigation method, at most two different methods, then report precise
588
+ supported navigation method, at most two different methods. A tool-required
589
+ interruption recovery may retry that same method once after fresh observation;
590
+ it is not a new method or an unlimited retry exemption. Then report precise
474
591
  `automation-blocked` and request the named panel once. User absence or an
475
592
  unanswered request never grants permission or proves the panel opened.
476
593
  App unavailable, restart/install, account/OAuth/paid actions outside explicit
477
- permission stay blocked. Do not repeat exhausted methods on ordinary continuation.
594
+ permission stay blocked; first check whether the needed action is already covered
595
+ by scoped authorization. Do not repeat exhausted methods on ordinary continuation
596
+ or repeat an unchanged blocker message.
478
597
  Record only fixed action/outcome codes, numeric times and version/source facts;
479
598
  never upload identities, credentials or conversation content. Unknown loaded
480
599
  version permits qualified observations only, not a version-specific fix claim.
@@ -536,7 +655,17 @@ Then run `pnpm verify:tarball -- artifacts/dsh-github-copilot-<package-version>.
536
655
 
537
656
  If a historical/approved local installation attempt is blocked solely by an enterprise registry or network policy while GitHub Actions can obtain unchanged pinned packages from an authorized official registry, record it as **local validation blocked**, not passed. Do not repeat dependency recovery, switch sources, or expand permissions. A PR may proceed with that limitation disclosed; the complete required GitHub CI matrix, including install, tests, build and package checks, is authoritative before merge. Never weaken or skip CI, substitute dependencies, bypass organizational restrictions, claim unrun local gates succeeded, or tag/release a revision whose required CI is not green. If CI has the same failure, resolve it through approved infrastructure rather than bypassing it.
538
657
 
539
- The release CI matrix targets only exact official `0.2.1-rc.3` on Windows and Linux; earlier pins are not support gates. It runs the unchanged tagged-source runtime fixture, optional native Chat fixture, exact published-artifact adapter fixture and all 28 native reviewer regressions without skips. This is a required gate, not evidence that an unexecuted qualification passed. `verify:upstream` is static seam-marker evidence. `verify:controlled-core` exclusively installs a temporary config fixture, refuses an existing target, and removes only its own file; it is not full plugin activation. Published-artifact synthetic adapter fixtures do not prove live provider transport. The release job must wait for the complete reusable CI matrix on the tagged revision, then verify its own packed bytes before publishing.
658
+ The release CI matrix targets only the exact current Core on Windows and Linux;
659
+ earlier pins are not support gates. It runs the unchanged tagged-source runtime
660
+ fixture, optional native Chat fixture, exact published-artifact adapter fixture
661
+ and the reviewer regressions required by the current CI workflow without skips.
662
+ This is a required gate, not evidence that an unexecuted qualification passed.
663
+ `verify:upstream` is static seam-marker evidence. `verify:controlled-core`
664
+ exclusively installs a temporary config fixture, refuses an existing target,
665
+ and removes only its own file; it is not full plugin activation. Published-artifact
666
+ synthetic adapter fixtures do not prove live provider transport. The release job
667
+ must wait for the complete reusable CI matrix on the tagged revision, then
668
+ verify its own packed bytes before publishing.
540
669
 
541
670
  ### Evidence and side effects
542
671
 
@@ -555,7 +684,7 @@ The release CI matrix targets only exact official `0.2.1-rc.3` on Windows and Li
555
684
  | Last metadata display | Same-account data may display during TTL refresh/loading/error but never authorize requests; credential/account/permission invalidation or proof expiry immediately revokes evidence; TTL does not extend tokens |
556
685
  | Last-success timestamp (alpha.7, #97) | Show `snapshot.discoveredAt` once beside count outside Manage: English relative text, full LOCAL date/time/zone tooltip and accessible semantic `time`; missing/invalid/no account hides it, future uses absolute text. Pending/error retains last success, successful refresh replaces it and sign-out clears it. One mounted 60s display-only timer makes no RPC/status/discovery calls and is disposed with timestamp/unmount; cache/discovery lifecycle unchanged |
557
686
  | Definitive `UNKNOWN_MODEL` | One bounded metadata refresh, never message replay or automatic model switching; generic HTTP/network errors are not guessed to be unknown models |
558
- | README screenshot | `copilot-search-routing.png` shows the current built plugin-detail routing card with synthetic providers in an isolated browser fixture. It proves only the current presentation surface, not live credentials, provider availability, search success or Desktop activation. |
687
+ | Historical routing screenshot | `copilot-search-routing.png` records a retired plugin-detail routing card with synthetic providers; it does not depict the current UI or prove live search or Desktop activation. |
559
688
  | Host attach/restart | Does not repair or rewrite the stored native profile; search proofs stay lazy and do not start at attach |
560
689
  | Background credential/reset notifications | Invalidate evidence and clear Client state with read-only status; no forced discovery on every token event; next Models open/use ensures metadata |
561
690
  | Eligible model/search request | May resolve/refresh credentials and run bounded capability proof; changes during proof must fail closed rather than reuse another account's proof |
@@ -565,6 +694,10 @@ Never say GPT-6/search works merely because settings, typecheck or a package imp
565
694
 
566
695
  ## Changing capabilities
567
696
 
697
+ The plugin-owned Auto and parent-follow implementations described in historical
698
+ paragraphs below are not mounted current behavior. Preserve retained data and
699
+ compatibility code, but do not reactivate their routing or UI.
700
+
568
701
  Auto routing, assessment, cost weights, continuity and assistance changes must
569
702
  first read and validate `docs/auto-iteration-review.json` using
570
703
  `scripts/iteration-review.mjs`. Follow `docs/evidence-driven-iteration.md`:
@@ -633,7 +766,7 @@ Retain native Host agent lookup/access checks and strict codecs. Failed reads
633
766
  must remain distinct from successfully read missing evidence; Retry reads the
634
767
  same turn only and never replays inference. See #249.
635
768
 
636
- Parent-model following uses the profile-wide `github-copilot.followParentModel` switch (default false), with a single control in the plugin-detail page (#234). Never require per-session enrollment for the ordinary experience. Preserve legacy `parentModelFollow` bindings independently, including when the broad switch is off. `src/parent-model-follow.ts` and the Auto Host integration use public projections and per-turn scoped routing, never descriptor/history/default writes. Child-owned explicit selections win; enabling authorizes replacement of creation-time route snapshots, not removal of later picker history. Only supported native spawn children and managed Copilot parents participate; roots, other providers, fork descriptors and historical dedicated policies remain native. Missing eligible parent evidence fails explicitly. Settings changes preserve admitted turns, and Client saves use one path-level CAS without credentials/discovery. See `docs/parent-model-follow.md` for requirements/mockup and `docs/automatic-model-routing.md` for exact-source evidence.
769
+ Historical parent-model following used the profile-wide `github-copilot.followParentModel` switch and a plugin-detail control (#234). That routing and control are retired; DSH owns current parent/child behavior. Preserve legacy `parentModelFollow` data without reactivating the plugin policy, mutating descriptors/history/defaults or silently migrating settings. See `docs/parent-model-follow.md` and `docs/automatic-model-routing.md` for historical evidence.
637
770
 
638
771
  Model preferences must survive the Client's owned-field projection before strict decoding. `excludedModelIds` is a hidden volatile Config leaf, read via `readInlineConfig`; exercise native SettingsForms CAS, fiber preservation and restart persistence, not only map-backed settings mocks. Missing preference settings leave model rows read-only with named diagnostics and status-only Retry. Unknown exclusion state is not an empty exclusion set. Discovery diagnostics stay in a separate Manage disclosure. Selected models are excludable without scanning Sessions/defaults; the legacy `lockedModelIds` wire leaf stays empty. Check exclusion at new-turn admission through public `agent/request`, preserving only the same Session/turn/model and exact native request signal for subsequent steps/retries. Clear admission on `turn/end`. Direct/unbound requests and new turns remain excluded; never rewrite selections/history or switch models silently. Account/token/metadata/cancellation guards remain independent and revocable.
639
772
 
@@ -645,7 +778,16 @@ Auto intent continuity (#293) uses the strict `githubCopilotAutoModelIntent` pub
645
778
 
646
779
  1. Identify the owning seam using `agent-contract.json`; do not duplicate an upstream owner.
647
780
  2. Add or update focused tests before changing deployment claims.
648
- 3. Update both READMEs when user behavior, setup, migration, or boundaries change.
781
+ 3. Review both READMEs and every affected domain document against source for
782
+ user behavior, setup, migration, limitations, and evidence changes. Update
783
+ both READMEs and the relevant `docs/*.md` in the same PR for user-facing
784
+ changes, beyond version-only edits. For unchanged behavior, explain why.
785
+ The PR `Documentation impact` declaration is checked against changed files
786
+ and rechecked when edited; it cannot judge semantic accuracy or relevance.
787
+ An agent or human reviewer must check those claims against source and retain
788
+ dated historical evidence. Record findings in the PR. This is not a new
789
+ human-only approval gate; agents may complete the review and proceed when
790
+ actual GitHub review/CI requirements and authorization boundaries are met.
649
791
  4. Update `deployment-baseline.json`, `agent-contract.json` and their verifiers when the corresponding contract changes.
650
792
  5. Build before package smoke; never hand-edit `lib/`.
651
793
 
package/CHANGELOG.md CHANGED
@@ -1,12 +1,29 @@
1
1
  # Changelog
2
2
 
3
- ## 0.1.0-alpha.13.dsh.0.2.2.rc.3 (prepared)
3
+ Status labels record what was known when each entry was written; immutable
4
+ release tags do not update this file after publication. Check the exact
5
+ [GitHub Release](https://github.com/cloga/dsh-github-copilot/releases) and npm
6
+ version separately rather than treating an old `prepared` label as current status.
7
+
8
+ ## 0.1.0-alpha.15.dsh.0.2.2.rc.3 (prepared)
9
+
10
+ - Include an optional, bounded offline multi-turn autonomy trace evaluator and synthetic examples (#494).
11
+ - Detect repeated approval/blocker requests, unconsented protected actions and unverified outcomes while preserving explicit task-scoped authorization across tool-state changes.
12
+ - Keep this as a repository analysis aid, not runtime agent enforcement or proof of improved autonomy; no profile, account or model calls are changed.
13
+
14
+ ## 0.1.0-alpha.14.dsh.0.2.2.rc.3
15
+
16
+ - Add independently enabled, content-free native Copilot logical stream and hosted-search aggregates (#486).
17
+ - Keep native observations separate from managed physical-request diagnostics; they do not prove HTTP dispatch, supplier receipt or execution.
18
+ - Preserve request, retry, Usage and settings ownership; old strict versions may not read retained new operation codes on rollback.
19
+
20
+ ## 0.1.0-alpha.13.dsh.0.2.2.rc.3 (published)
4
21
 
5
22
  - Qualify native Copilot profiles after actual Config/Settings normalization: schema-default empty template maps and boolean `supportsStrictMode` no longer disable reviewer-only temperature omission or canonical hosted-search readiness (#483).
6
23
  - Apply the same narrow compatibility rule at provider, model and model-override levels while retaining unknown-value, credential, endpoint, headers, ownership and lifecycle refusal.
7
24
  - Add normalized native configuration and reviewer/search regressions. No profile rewrite, Core patch, automatic restart, general sampling override or live supplier acceptance is implied.
8
25
 
9
- ## 0.1.0-alpha.12.dsh.0.2.2.rc.3 (prepared)
26
+ ## 0.1.0-alpha.12.dsh.0.2.2.rc.3 (published)
10
27
 
11
28
  - Retire the managed route's projected-image MIME preflight and Goal schema rewriting (#481). Core owns image projection and Goal schemas; supplier errors are not preempted by a plugin MIME diagnostic.
12
29
  - Keep the managed route's permission-field filtering and the native model-sync format eligibility checks unchanged.
package/CONTRIBUTING.md CHANGED
@@ -16,9 +16,9 @@ Core inspection is read-only. Isolated tests may use unchanged pinned Core artif
16
16
 
17
17
  1. Determine `pwd`, Git status, remote and actual default branch. Preserve other people's changes. Start a feature branch such as `feature/issue-73-agent-readiness`; branch names need not contain an owner name.
18
18
  2. Read `node scripts/agent.mjs describe --json` and run `node scripts/agent.mjs doctor --json`. These are read-only repository diagnostics, not live health checks. Plans from `node scripts/agent.mjs plan <task> --json` contain unexecuted argv lists.
19
- 3. Use Node 24 LTS for development (runtime floor: 22.19.0), the pinned pnpm version, and frozen install inside the checkout/worktree.
20
- 4. Add focused regression tests before changing capability claims. Keep READMEs, deployment evidence and the Agent task map aligned.
21
- 5. State expected outcomes and scope; run:
19
+ 3. Use Node 24 LTS for development (runtime floor: 22.19.0) and the pinned pnpm version. Agents must follow the [environment guardrails](./docs/environment-evidence-guardrails.md): local work is limited to the guarded dependency-free scripts tests; do not install or restore dependencies locally.
20
+ 4. Add focused regression tests before changing capability claims. Review and update both READMEs and relevant domain docs with user-facing behavior; keep deployment evidence and the Agent task map aligned.
21
+ 5. State expected outcomes and scope. The existing GitHub Actions workflow runs:
22
22
 
23
23
  ```sh
24
24
  pnpm install --frozen-lockfile
@@ -26,10 +26,10 @@ Core inspection is read-only. Isolated tests may use unchanged pinned Core artif
26
26
  pnpm pack --pack-destination artifacts
27
27
  ```
28
28
 
29
- Validate the resulting exact path with `pnpm verify:tarball -- artifacts/dsh-github-copilot-<package-version>.tgz`. The release task plan supplies that path automatically. Never run broad cleanup on an existing artifacts directory just to package again.
29
+ CI validates the resulting exact path with `pnpm verify:tarball -- artifacts/dsh-github-copilot-<package-version>.tgz`. The release task plan supplies that path automatically. Never run broad cleanup on an existing artifacts directory just to package again. Local agent validation uses `node scripts/agent.mjs guard --task tooling --environment local-auxiliary --operation scripts-test`; it is not the complete gate.
30
30
  6. Inspect the complete diff, compare expected versus actual results, and report what was NOT checked. Source markers, stubbed unit tests, built module imports, upstream config fixtures, authenticated API calls, installation, and loaded UI are different evidence layers.
31
31
  7. Commit with the issue reference and accurate attribution. For work assisted by DSH use `Assisted-by: DeepSeek Harness (DSH)`. Other agents must name the actual tool; do not infer authorship from the model vendor or copy a bot identity. Add `Co-authored-by` only for real verified collaborators. No fabricated email, author replacement or published-history rewrite.
32
- 8. Push the feature branch and open a PR containing `Fixes #<issue>`, verification, risks and rollback. For user-requested delivery, merge without an additional approval prompt after required review/CI and branch protection; respect explicit code-only, review-only or do-not-merge restrictions.
32
+ 8. Push the feature branch and open a PR containing `Fixes #<issue>`, verification, risks and rollback. For runtime, tooling, release-metadata, workflow or agent-contract changes, fill the PR's **Documentation impact** section with `Behavior: user-facing` or `Behavior: unchanged` and an explanation. User-facing changes must list substantively changed `README.md`, `README.zh.md` and a relevant `docs/*.md`; version-only updates are insufficient. The CI gate rechecks body edits and only checks structure/file differences, not whether the assessment is true or complete. An agent or human reviewer must check affected behavior and evidence against all relevant docs and record findings in the PR; this adds no human-only approval gate. For user-requested delivery, merge without an additional approval prompt after actual required review/CI and branch protection; respect explicit code-only, review-only or do-not-merge restrictions.
33
33
 
34
34
  Generated `lib/` output, archives, `.env`, tokens, local credentials, machine state and production screenshots with real authorization codes must not be committed. Use synthetic fixtures for screenshots and report staged UI honestly.
35
35
 
package/README.md CHANGED
@@ -50,6 +50,11 @@ confirm that the exact version has passed the required gates and its Release
50
50
  and npm artifacts are available. Published, installed, and loaded states are
51
51
  distinct.
52
52
 
53
+ Plugin `0.1.0-alpha.14.dsh.0.2.2.rc.3` passed the Windows/Linux source and
54
+ published-artifact [CI](https://github.com/cloga/dsh-github-copilot/actions/runs/43691055529)
55
+ and was published to GitHub Releases and npm. This does not verify Desktop
56
+ artifacts, installation, loaded state or live supplier calls.
57
+
53
58
  Use the read-only exact-Core selector from a verified source checkout to check
54
59
  the release mapping:
55
60
 
@@ -57,8 +62,9 @@ the release mapping:
57
62
  node scripts/select-core-release.mjs 0.2.2-rc.3
58
63
  ```
59
64
 
60
- It must report a fully verified published release before an install is
61
- recommended. See [release selection and distribution](./docs/npm-distribution.md)
65
+ The selector checks Release/npm metadata and named assets; it does not run CI,
66
+ hash downloaded bytes, or verify a loaded Desktop. Check those separately before
67
+ recommending installation. See [release selection and distribution](./docs/npm-distribution.md)
62
68
  and the [model compatibility acceptance criteria](./docs/model-compatibility-acceptance.md).
63
69
 
64
70
  ## Use
@@ -67,15 +73,15 @@ After the exact compatible release is available and verified, download its
67
73
  original asset and checksum in an empty directory:
68
74
 
69
75
  ```sh
70
- curl -fLO https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.13.dsh.0.2.2.rc.3/dsh-github-copilot-0.1.0-alpha.13.dsh.0.2.2.rc.3.tgz
71
- curl -fLO https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.13.dsh.0.2.2.rc.3/SHA256SUMS
76
+ curl -fLO https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.15.dsh.0.2.2.rc.3/dsh-github-copilot-0.1.0-alpha.15.dsh.0.2.2.rc.3.tgz
77
+ curl -fLO https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.15.dsh.0.2.2.rc.3/SHA256SUMS
72
78
  sha256sum --check SHA256SUMS
73
79
  ```
74
80
 
75
81
  For the `web` profile, install the same Release asset:
76
82
 
77
83
  ```sh
78
- dsh plugin --profile web add https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.13.dsh.0.2.2.rc.3/dsh-github-copilot-0.1.0-alpha.13.dsh.0.2.2.rc.3.tgz
84
+ dsh plugin --profile web add https://github.com/cloga/dsh-github-copilot/releases/download/v0.1.0-alpha.15.dsh.0.2.2.rc.3/dsh-github-copilot-0.1.0-alpha.15.dsh.0.2.2.rc.3.tgz
79
85
  ```
80
86
 
81
87
  Choose the intended profile explicitly. Then:
@@ -147,8 +153,8 @@ another model. Each search makes one fixed capability probe and, only after
147
153
  proof, one user query; both may incur charges.
148
154
  Custom endpoint, credential or transport overrides and unproven models remain
149
155
  unavailable. There is no model retry, backend replacement or plugin routing UI.
150
- The native Auto reviewer's qualified Copilot Responses requests omit
151
- `temperature`; ordinary Chat and native allow/deny/error/cancel behavior remain
156
+ When qualification succeeds, the native Auto reviewer's Copilot Responses requests
157
+ omit `temperature`; ordinary Chat and native allow/deny/error/cancel behavior remain
152
158
  unchanged. Qualification uses the native adapter's pi-ai catalog even when
153
159
  the plugin resolves a separate copy; missing protocol facts remain unavailable.
154
160
  Native schema-default empty `chatTemplateKwargs`/`chatTemplateArgs` maps and a
@@ -157,12 +163,35 @@ configuration. Other compatibility overrides remain unqualified. No settings
157
163
  rewrite is needed to remove schema defaults; do not disable review or search
158
164
  proof as a workaround. CI acceptance and live Desktop activation are separate.
159
165
  See [reviewer compatibility](./docs/auto-review-sampling.md).
166
+
167
+ Plugin releases through alpha.12 can reject DSH's schema-default compatibility
168
+ maps and disable both enhancements ([#483](https://github.com/cloga/dsh-github-copilot/issues/483)).
169
+ The alpha.13 release fixes that qualification; live runtime behavior still
170
+ requires separate verification. Registration or model metadata alone proves
171
+ neither works.
172
+
160
173
  The native route's quota and optional identity display use only the canonical OAuth account. If the
161
174
  provider names a credential, endpoint or nonempty headers override, or its configuration
162
175
  cannot be verified, quota is shown as unavailable rather than attributed to a
163
176
  different account. This is not proof of request-level billing attribution.
164
177
  Native context-incident diagnostics remain available without the optional
165
178
  managed continuation service; they do not replace DSH's context meter.
179
+ Local diagnostics also offers separately opt-in, content-free logical
180
+ observations for native Copilot streams and the optional Copilot search backend.
181
+ They do not expose HTTP details or prove supplier receipt or execution.
182
+ See [local diagnostics](./docs/plugin-diagnostics.md).
183
+
184
+ ### Other optional capabilities
185
+
186
+ - On the managed route, Credits can select an already saved account for a
187
+ Session's subsequent turns. [Visible-history continuation](./docs/session-continuation.md)
188
+ is a disclosed, lossy Responses policy; existing Sessions are not silently
189
+ enrolled, and enabling it never sends or retries a message.
190
+ - [Segmented compaction recovery](./docs/manual-compaction-recovery.md) requires
191
+ explicit engine selection; installing the plugin does not replace Core's engine.
192
+ - [Local diagnostics](./docs/plugin-diagnostics.md) and managed Responses gzip
193
+ are default-off; enabling diagnostics does not upload data. GitHub Enterprise
194
+ Cloud tenant login is supported, while GHES and Enterprise quota are not.
166
195
 
167
196
  ### Removing the enhancement
168
197
 
@@ -198,6 +227,11 @@ The exact development target is in `deployment-baseline.json`; follow
198
227
  and package/release gates run in GitHub Actions. A successful local source
199
228
  check, synthetic fixture, release publication, profile installation, and loaded
200
229
  Desktop runtime are different evidence layers; report each accurately.
230
+ The optional [offline autonomy evaluator](./docs/agent-autonomy-evaluation.md)
231
+ analyzes explicitly supplied sanitized traces; it does not run or control agents.
232
+ PRs declare documentation impact; agent or human reviewers still verify semantic
233
+ accuracy and coverage. This adds no human-only approval gate. See
234
+ [contributing](./CONTRIBUTING.md#change-workflow).
201
235
  Commit attribution must use an `Assisted-by` trailer for the actual assisting
202
236
  tool; do not invent a co-author identity.
203
237