Why: Both supported tool finish reasons must fail closed when no call was collected; the Dev composition pins that exact Agent-Native revision through Agent-Squad and Backend.
Evidence: The function_call regression failed before the fix and passed afterward. Agent-Native, Agent-Squad and Backend full vet/test suites, strict Dev pins, and task-owned PostgreSQL/browser checks passed at their recorded versions. Current Dev HEAD CI and final independent replay remain pending.
Impact: Component gitlinks and task evidence only; no paid model call or deployment.
Refs: U3 r0.4 review Important 2; Agent-Native PR #46; Agent-Squad PR #53; Backend PR #193
Why: Close the independent review blockers and keep Dev, Backend, Agent-Squad and Agent-Native on one exact stream/runtime combination.
Evidence: Review issues reproduced RED then passed; component suites and six HTTP/Playwright cases passed. The staged Dev composition passed make test && make build with 340 governance tests and strict pins; full original receipts are in the task card. Current committed-head CI and review closure remain pending.
Impact: Component gitlinks and task evidence only; no paid A01 or deployment.
Refs: U3 r0.4 Task 6; Agent-Native PR #46; Agent-Squad PR #53; Backend PR #193; Frontend PR #79
Why: Fix the exact Agent-Native consumer chain and publish the r0.4 Backend and browser test candidate as a reviewable Dev combination.
Evidence: Strict pin preflight, component tests, isolated PostgreSQL, six HTTP/Playwright cases, and make build passed. The full local make test was interrupted by an unrelated concurrent fixed-port collision; this failure and prior pin failures remain in the task card. Exact-HEAD CI and independent review remain pending.
Impact: Dev gitlinks and task evidence only; no paid model call or deployment.
Refs: U3 r0.4 implementation plan; Agent-Native PR #46; Agent-Squad PR #53; Backend PR #193; Frontend PR #79
Why: The approved document dependencies have been reviewed and merged. Their actual immutable merge commits must be fixed before final host validation, while operational activation remains a separate gate.
Evidence: Independent document replay and verified PR merges are recorded in the task card. The staged host combination passed make test with actual merged child pointers; final commit build/CI remains pending and will be recorded outside source.
Impact: Advance only the two reviewed merged child pins, expose the real plan path, and record partial stage-seven approval and pending activation. Preserve current main components and existing admission/owner boundaries.
Refs: Guidelines PR #5 (2f5b198); Documents PR #96 (129cc35); Dev PR #519; implementation plan U4-U6; Issue #518
Why: backend main moved to 017b40f (agent-squad pin bump, E1 chain #129) and
Dev's top-level agent-squad must match the nested pin the components record,
otherwise test-pins flags container/Dev divergence.
Evidence: check-pins green and make test rerun recorded on this head.
Impact: modules/backend 5714c73 -> 017b40f; deps/agent-squad 8bbf850 -> 490b0b2.
Refs: Backend PR #128/#129; agent-squad PR #51.
Why: the backend pin 5714c73 records agent-native c348441 (E1 parallel-tools
kernel); Dev's top-level agent-native gitlink must move with it or test-pins
flags the container/Dev divergence (three-layer pin cascade).
Evidence: check-pins green after the bump (recorded via make test evidence).
Impact: deps/agent-native gitlink 3b92cb6 -> c348441 only.
Refs: Backend PR #128; agent-native PR #44.
Why: G2-2 lands plugin slash aliases (EXT-10②) and the intent
declaration surface — the bus binds plugin-spec Slash fields into the
shared alias table (four entrances, one audit envelope), extensions
declare intents via AddIntent, and the system.intent builtin routes
natural language through the bus deterministically (longest-phrase
containment). The umbrella carries the backend gitlink forward.
Evidence: Backend #122 head 9c4edc4 CI success, merged as 9a531a6;
full backend suite 32/32 packages green; lib pin Agent-Command 4f583ca
(PR #6).
Impact: umbrella gitlink only.
Refs: plan G2-2; Backend #122; Agent-Command #6.
Why: consuming backend main (#115 kernel-transcript + #116 EXT-14
dependency inversion) moves backend pins for agent-native (3b92cb63)
and agent-squad (8bbf850, Squad #49 nested native bump); the umbrella
check-pins requires Dev's top-level pins and every container gitlink to
agree, and Dev's agent-command pin must follow backend's EXT-14 lib bump
(5c6dd0c, lib PRs #4/#5).
Evidence: Backend #116 (6337cd8) and #117 (5544a03) vet-and-test CI
success, merged as 8157e68/31965fd; Squad #49 merge 8bbf850 records
deps/agent-native=3b92cb63 (ls-tree verified); umbrella
lint-boundaries pre-verified green at 6337cd8 (77 edges, negative probe
ok); full make test recorded on the task card at the final head.
Impact: gitlink-only changes (modules/backend, deps/agent-squad,
deps/agent-command, deps/agent-native); no source edits in the umbrella.
Refs: Backend #115/#116/#117; Squad #49; agent-command lib #4/#5;
proposal 2026-10-09-ext-g1-b1-b3 (EXT-14).
Why: check-pins requires nested and Dev-level pins to agree; squad main
advanced to 819b0c5 (nested agent-native pin #48) and backend main to
fe7c245 (nested agent-squad pin #103) after the Q3 WP1-WP5 merges.
Evidence: both component PRs CI green (GATE); make test recorded via
make evidence in the follow-up commit on this branch.
Impact: deps/agent-squad 663e89e -> 819b0c5; modules/backend 4791b2e ->
fe7c245; deps/agent-native stays 37c5e6f.
Refs: Q3 plan WP6; task card .task/feature-agent-stack-q3.md
Why: Q3 (convergence and interop) completed across agent-native (#40),
agent-squad (#46/#47) and backend (#99-#102); Dev gitlinks must track the
merged component mains and the task card records the outcome before the
completion declaration.
Evidence: card increment table rows WP1a-WP6 with merged PR/SHA references;
component CI statuses recorded as GATE in the card verification section;
three-segment outputs (precheck/task-sync/test) appended via make evidence
in follow-up commits on this branch.
Impact: deps/agent-squad bcc8d08 -> 663e89e, deps/agent-native 1ac21c6 ->
37c5e6f, modules/backend 06934bc -> 4791b2e; task card filled.
Refs: .task/feature-agent-stack-q3.md; Q3 plan 2026-10-07-agent-stack-q2q3-plans.md
Why: Q2 (topology library) completed in Agent-Squad across WP1-4 (PRs
#42-#45, main bcc8d08) and the backend nested pin followed (PR #98, main
06934bc); Dev gitlinks must track the merged component mains and the task
card records the outcome before the completion declaration.
Evidence: card increment table rows WP1-4 with merged PR/SHA references;
component CI statuses recorded as GATE in the card's verification section;
three-segment outputs (precheck/task-sync/test) appended via make evidence
in follow-up commits on this branch.
Impact: deps/agent-squad c462bc4 -> bcc8d08, modules/backend 6e02a1f ->
06934bc; task card filled (results/verification/increments).
Refs: .task/feature-agent-stack-q2.md; Q2 plan 2026-10-07-agent-stack-q2q3-plans.md
Why: Backend#96 merged the usage split and default retry; the core
Usage fields, Frontend#48, and the native/squad cascades landed; Dev
pins and the increment record follow.
Evidence: check-pins pins ok; WP6 chain CI green on every merge head;
audit N20 closes (split flows into run records on every path).
Impact: usage accounting is uniform end to end; stream retry is a
library default.
Refs: Agent-Native#35-#39, Agent-Squad#41, Agent-System-Backend#96,
Agent-System-Frontend#48
Why: Backend#95 merged the memory port wiring; Native#34 and Squad#38
carried the port; Dev pins and the increment record follow.
Evidence: check-pins pins ok; WP5 chain CI green on every merge head;
the migration parity test GATE on backend 766a96e8.
Impact: the memory port exists end to end with write-back.
Refs: Agent-Native#34, Agent-Squad#38, Agent-System-Backend#95
Why: Backend#94 merged the checkpoint closure wiring; Native#33 and
Squad#37 carried the library work; Dev pins and the increment record
follow.
Evidence: check-pins pins ok; WP3 chain CI green on every merge head;
acceptance clause 2's tests GATE across layers.
Impact: rollback evidence complete (eligibility, symlink guard,
partial restore, retention, HTTP).
Refs: Agent-Native#33, Agent-Squad#37, Agent-System-Backend#94
Why: Backend#93 merged the GenAI telemetry wiring; Native#32 and
Squad#36 carried the convention work; Dev pins and the increment
record follow.
Evidence: check-pins pins ok; WP4 chain CI green on every merge head;
acceptance clause 3's tests GATE on native fe0c4a9f.
Impact: spans follow the OTel GenAI convention end to end.
Refs: Agent-Native#32, Agent-Squad#36, Agent-System-Backend#93
Why: Backend#92 merged the JournalOnly conversation wiring; Native#31
and Squad#35 carried the library mode; Dev pins and the increment
record follow, including the design adjustment (journal-only instead
of transcript-authoritative, with the freeze rationale).
Evidence: check-pins pins ok; WP2 chain CI green on every merge head;
acceptance clause 1's product tests GATE on backend 9f3fdb26.
Impact: conversation runs are durable mid-turn without freezing the
context pipeline.
Refs: Agent-Native#31, Agent-Squad#35, Agent-System-Backend#92
Why: Backend#91 merged the graceful-resume split; Native#30 and
Squad#34 carried the tail repair and N6 fix; Dev pins and the
increment record follow.
Evidence: check-pins pins ok; WP1 chain CI green on every merge head;
acceptance clause 1's library tests GATE on native cb21ec4e.
Impact: every Run now pairs orphan tails before the first request;
gracefully-interrupted subagents resume exactly once.
Refs: Agent-Native#30, Agent-Squad#34, Agent-System-Backend#91
Why: Backend#90 merged the subagent transcript/resume/notice work;
Native#29 and Squad#33 carried the Background field; Dev pins and the
increment record follow.
Evidence: check-pins pins ok; WP5 chain CI green on every merge head;
the three WP5 acceptance tests are GATE on backend main 7980a8a2.
Impact: subagents are resumable with per-run transcripts and durable
notices.
Refs: Agent-Native#29, Agent-Squad#33, Agent-System-Backend#90
Why: Backend#89 merged the plugin/bundled skill validation; Native#28
and Squad#32 carried the validator; Dev pins, the plugins→skills edge,
and the increment record follow.
Evidence: check-pins pins ok; WP4 chain CI green on every merge head;
TestV2InvalidSkillRejected GATE on backend main b44f78b0.
Impact: skills now validated against the open Agent Skills spec.
Refs: Agent-Native#28, Agent-Squad#32, Agent-System-Backend#89
Why: Native#27 merged the MCP 2026-07-28 alignment; Squad#31 and
Backend#88 carried the cascade; Dev pins and the increment record follow.
Evidence: check-pins pins ok; WP3 chain CI green on every merge head;
TestStdioSlowCancelThenEcho (acceptance 3) passes unchanged.
Impact: MCP client negotiates versions and recognizes resultType.
Refs: Agent-Native#27, Agent-Squad#31, Agent-System-Backend#88
Why: Native#26 merged the native Anthropic client (acceptance clause 1,
audit N4 closed); Squad#30 and Backend#87 carried the cascade; Dev pins
and the increment record follow.
Evidence: check-pins pins ok; WP2 chain CI green on every merge head;
TestAnthropicPresetToolRound GATE on native main d6053baf.
Impact: acceptance clause 1 is closed.
Refs: Agent-Native#26, Agent-Squad#30, Agent-System-Backend#87
Why: Backend #86 merged the tool-group-aligned trimming; native #25 and
squad #29 carry the spill and cache contract work; the Dev pins and the
increment record follow.
Evidence: check-pins pins ok; WP1 chain CI green on every merge head.
Impact: acceptance clause 2's product side is closed
(TestTrimOldestKeepsToolPairs GATE).
Refs: Agent-Native#25, Agent-Squad#29, Agent-System-Backend#86
Why: native ead690be (Agent-Native#24) stops workspace-less requests
from writing session state into the process CWD; squad 029edc08
(Agent-Squad#28) and backend 68addfda (Backend#85) carry it. Without
this the CI build step fails the worktree check after make test.
Evidence: check-pins pins ok; combination re-verified by make test
before push (see card).
Impact: none beyond pins; the I2 slow-lane stack is unchanged.
Refs: Agent-Native#24, Agent-Squad#28, Agent-System-Backend#85, Dev#242
Why: backend #83 merged the slow-lane wiring (intentsem scorer/
generator, gate SetSemantic/SetGenerator, settings flags default off);
the Dev pins and boundary rules must move with it.
Evidence: backend CI green on 9d29371; check-pins consistent
(agent-intent 343b1e6 across Dev and the backend nested copy);
lint-boundaries registers intentsem as an agent-intent adapter and the
architecture baseline gains backend/app->backend/intentsem.
Impact: product behavior unchanged by default; the two
runtime.intent* flags opt into the slow lane.
Refs: Agent-System-Backend#83, I2 increment 3
Why: Agent-Squad PR #26 and Backend PR #82 landed; Dev pins must
point at those merged component main heads.
Evidence: check-pins exit=0.
Impact: Two gitlinks.
Refs: Agent-Squad PR #26; Agent-System-Backend PR #82
Why: Agent-Squad PR #25 and Backend PR #81 landed; Dev pins must
point at those merged component main heads.
Evidence: check-pins exit=0.
Impact: Two gitlinks.
Refs: Agent-Squad PR #25; Agent-System-Backend PR #81
Why: Agent-Squad PR #24 and Backend PR #80 landed; Dev pins must
point at those merged component main heads.
Evidence: check-pins exit=0.
Impact: Two gitlinks.
Refs: Agent-Squad PR #24; Agent-System-Backend PR #80
Why: Agent-Squad PR #23 and Backend PR #79 landed; Dev pins must
point at those merged component main heads.
Evidence: check-pins exit=0.
Impact: Two gitlinks.
Refs: Agent-Squad PR #23; Agent-System-Backend PR #79
Why: The I1 eval increments landed across the chain (intent #8,
squad #21/#22, native #22, backend #77/#78); Dev pins must point at
those merged component main heads.
Evidence: check-pins exit=0 after full matrix propagation.
Impact: Four gitlinks.
Refs: Agent-Intent PR #8; Agent-Squad PR #21/#22; Agent-Native PR #22; Agent-System-Backend PR #77/#78
Why: The I1 intent/router increments landed across the chain (intent
#6/#7, squad #19/#20, native #21, backend #74/#75); Dev pins must
point at those merged component main heads.
Evidence: check-pins exit=0 after the full cascade.
Impact: Four gitlinks.
Refs: Agent-Intent PR #6/#7; Agent-Squad PR #19/#20; Agent-Native PR #21; Agent-System-Backend PR #74/#75
Why: Increment four landed across the chain (native #19/#20, backend
#72, squad #18, backend #73); Dev pins must point at those merged
component main heads.
Evidence: check-pins exit=0; native c9c2f23, squad 7ecc04e and
backend 303983bd are direct merge commits on their component mains.
Impact: Three gitlinks.
Refs: Agent-Native PR #19/#20; Agent-Squad PR #18; Agent-System-Backend PR #72/#73
Why: Increment three landed across the chain (native #18, squad #17,
backend #71); Dev pins must point at those merged component main
heads so the tested combination is the released one.
Evidence: check-pins exit=0; native c5b9dca, squad fc9620e and
backend f0b0db9 are direct merge commits on their component mains.
Impact: Three gitlinks.
Refs: Agent-Native PR #18; Agent-Squad PR #17; Agent-System-Backend PR #71
Why: Squad PR #16 and Backend PR #70 completed the pin cascade;
Dev's squad and backend pins must point at those merged heads for
check-pins to verify the tested combination.
Evidence: a75d359 is the #16 merge on squad main; backend main head
is the #70 merge; backend CI passed before each merge.
Impact: Two gitlinks.
Refs: Agent-Squad PR #16; Agent-System-Backend PR #70
Why: Library PR #17 and Backend PR #69 are merged; the Dev pins must
point at the merged component main heads (native c2befaf, backend
fe0c5a4) so the tested combination is the released one.
Evidence: both pins are direct merge commits on their component
mains; backend CI passed at 4075cc6 before its merge.
Impact: Two gitlinks plus the task card evidence records.
Refs: Agent-Native PR #17; Agent-System-Backend PR #69
Why: declarations were passing as capability; hand-transcribed outputs flowed
through cards and PR bodies as if they were execution evidence, and stale
external facts (e.g. the dead git-panel host) kept being trusted until burned.
Evidence: guidelines draft.10 (PR Development-Guidelines#4, merged af199ee)
pinned here; policy/facts.md F1-F4 each re-verified on 2026-10-05 before
registration (endpoint version, token HTTP 200 via header only, SSH ls-remote,
dead-host negative probe); scripts syntax-checked (bash -n); tool self-tests
and make test recorded in the task card via the new make evidence.
Impact: new AGENTS.md section (SELF/REPLAY/GATE, tool-written evidence,
replay-on-review, zero-trust facts, honest-empty); make evidence/facts-verify;
precheck gains [hooks]/[facts], task-sync gains stale-evidence-HEAD warning;
task-start cards now embed the evidence table; gitlink advances to draft.10.
Refs: task card .task/feature-evidence-driven-gates.md; guidelines CHANGELOG
0.1.0-draft.10.
Why: Dev's agent-native moved for message shaping, so squad and backend have to follow or check-pins fails.
Evidence: Squad 5a57636 and backend include that pin. make test previously stopped at test-pins on this mismatch.
Impact: The parent, backend, and squad all resolve agent-native 3a110a7.
Refs: .task/feature-sillytavern-review.md
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: The research record, HTML cleaning, listen guard, secret store, and context controls land together, and the parent pointers have to name those commits.
Evidence: Child commits are core bd63223, agent-native 3a110a7, ui d76a8e6, frontend 36b92ed, documents f3f823e, backend 42cca8f. svc.sh forwards the listen switches and info prints them.
Impact: A reverse proxy or a LAN hostname needs allowedHosts or an unauthenticated host check returns 403. Startup moves raw API keys into the secret table after a settings snapshot.
Refs: .task/feature-sillytavern-review.md
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: the shared Python host and the MCP connect change landed in the libraries the product already calls.
Evidence: native, squad, and backend feature commits are recorded on the task card.
Impact: gitlinks move to those commits. Code orchestration stays off unless the setting is turned on.
Refs: feature/code-orchestration
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: tool intent, file checkpoints, fallback, and span names landed in the libraries and the product shell.
Evidence: native, squad, core, backend, frontend, and ui feature commits are recorded in the task card.
Impact: gitlinks move to those commits. Default fallback stays empty and replay stays never.
Refs: feature/durable-rewind
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Session mode, confirmation detail, and the shared tool gate now exist in the libraries the product builds.
Evidence: make test passed. Plan mode rejects write_artifact while the default mode still runs it. Hook adapters match the old stop, reminder, and deny counts.
Impact: Empty permission mode keeps today's profile policy. There is no mode that skips every confirmation.
Refs: feature/permission-hooks
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Tool-result clearing, folded reminders, and character truncation land in the native loop and the product path together.
Evidence: agent-native 6fd1d44, agent-squad 4d53450, backend b2dc9a7.
Impact: Default context window stays 32000 characters. Empty tool selection still exposes the profile's tools.
Refs: feature/context-engineering
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: The direct path's default retry only works once every repo builds the same model stream.
Evidence: native 9f5751c, squad 9efce79, backend 93adb93.
Impact: Three gitlinks move. No product UI change.
Refs: feature/llm-robustness
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Dev has to record the native head that treats a cancelled stream as aborted, and the squad and backend checkouts that point at it.
Evidence: GOWORK=off go test -count=50 -race ./runtime/llm/ passed on d84a67c.
Impact: gitlinks only. The earlier make test covered the previous pin.
Refs: D1
Why: Dev has to test the split modules and refuse a core import of a sibling library.
Evidence: bash scripts/lint-boundaries.sh printed boundaries ok. Full make test follows on this pin.
Impact: test-libs runs the intent, native, sqlite, and command modules. The sibling allowlist is empty.
Refs: D1
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: native 855fe41 is the first head whose go.sum covers intent's NFKC import, so the later pins have to move with it.
Evidence: native go test of runtime, adapters, and exec passed; squad and backend vet passed on 855fe41 and 31ee395.
Impact: three gitlinks and the task card move. Product code in Dev is unchanged.
Refs: I4
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Dev must record the intent, native, squad, and backend commits that close this batch so check-pins sees one tree.
Evidence: agent-intent 05e74d8, agent-native 11d1a76, agent-squad 1bb0c3a, and backend a494375 are the feature/agent-stack-r2 heads.
Impact: four gitlinks move and the task card records the batch scope.
Refs: R2 pin
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Dev must record the commits that contain the hold fixes, the offline pin check, and the CI checkout change.
Evidence: intent 7f7879b06b053f0a62746309b79e0e1b9a00f42d, native 91c155618487e21edeb4f1d3cc27e985aae2b0d2, squad e928ea3c0fa649cbbbad22a93a1bc0a12dee5f7d, backend ef0a1bef0f7bee626e59f160be51fb0e5f6aa692.
Impact: The four gitlinks match the open pull request heads. Other pins are unchanged.
Refs: feature/agent-stack-r0-r1
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Dev could ship a combination whose nested gitlinks differed from the top-level pins, and make test never ran the three agent libraries.
Evidence: A go.work was measured and not adopted: workspace MVS selected newer golang.org/x crypto, exp, net, term, and text than backend with GOWORK=off. check-pins passes on the new pins and fails when backend's squad pin is 5cbac00.
Impact: make test adds pin checks and GOWORK=off vet/test for intent, native, and squad. make build still compiles inside backend; the pin check requires those gitlinks to match Dev.
Refs: R0
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: Tool calls stayed invisible while arguments streamed, and a failed run hid the only reply inside the collapsed process block.
Evidence: make test in the worktree passed: boundaries, core, backend, frontend vitest 66, ui vitest 79, agent-command. Mock browser check showed the Chinese process summary, the reply outside the fold, a humanized error with 继续执行, and the back-to-bottom button.
Impact: Dev gitlinks move to the feature commits for core, agent-native, backend, ui, and frontend. DESIGN.md §9.6 describes individual tool blocks for exec, write, and delegate.
Refs: feature/stream-display-overhaul
Co-authored-by: Cursor <cursoragent@cursor.com>
Why: the direct chat path and both consumers must compile the same agent-native commit, or the new loop fields exist in only one replace target.
Evidence: make test passed (boundaries, core, backend, frontend 51, ui 71, agent-command). The three agent-native gitlinks are a34689c.
Impact: Dev pins deps/agent-native, deps/agent-squad, and modules/backend. Squad behavior stays on the zero-value kernel switches.
Refs: feature/agent-harness-p1
Co-authored-by: Cursor <cursoragent@cursor.com>
weichen
