Commit Graph

1451 Commits

Author SHA1 Message Date
kingwl
c6ba5f1d26 Merge origin/master (agent-scope runtime #224, session-query, run-gates CI harness)
Conflict resolution beyond line merges: the ACP bridge's mode/set
re-notify moved inside master's new containment finally (a throwing
presenter must not desync the picker), and session/new keeps the modes
advertisement beside master's plain object return. dsh-mode's boundary
flush moved off the session/event feed onto the loop's interception
seams (agent/prompt-submit / agent/turn-continuation): master's scoped
dispatch made post-commit session observers observe-only — an append
from one re-enters the publishing append — while the two waterfalls
fire outside any log emit and bracket the same boundaries; mode.spec's
boundary helper now dispatches those seams via agentEvents, and the
containment tests stub only the mode/set append. Catalogs, doc graphs,
and the module graph regenerated.
2026-07-13 15:12:11 +08:00
Tianyi Cui
b3e1a4154b Merge pull request #224 from deepseek-harness/worktree-agent-scope-design
Agent-scoped registrations: composition, events, and lifecycle
2026-07-13 14:10:45 +08:00
Tianyi Cui
6a8118c967 docs: align scoped runtime contracts 2026-07-13 13:41:27 +08:00
Tianyi Cui
e04ec07345 refactor(core): remove owner-final assembly machinery 2026-07-13 13:09:41 +08:00
Tianyi Cui
f32cfafa1a refactor: narrow synchronous extension contracts 2026-07-13 11:58:55 +08:00
Tianyi Cui
f7b9cea733 docs: clarify scoped persona shadowing 2026-07-12 23:40:27 +08:00
Tianyi Cui
5d49a79165 docs: align lifecycle guides with final entry 2026-07-12 23:34:35 +08:00
kingwl
1fe2f99580 refactor(mode): drop the per-mode tool allowlist — enforce where an enforcer exists
A ModeDefinition is now exactly { section, access? }; unknown keys (a
tools list included) fail loud at load. What plan mode still does: the
guidance section, the exit_plan_mode visibility rule (plan only, both
soft surfaces), the access cap's bash/resolve-mode clamp, and the two
cap-derived pre-execute guards (the bash trio is withheld when no
confining executor can honor the cap; sandbox escalation is denied
while it holds). The general deny-by-default gate and the assemble
allowlist filter are gone: which tools a mode admits is an effects
question, and a hand-maintained name list mislabels it — it must track
every composed tool and rots silently as tools arrive. The dimension
returns as a consumer of effects self-declaration on tool definitions
(MCP ToolAnnotations as the template) — rationale and restart trigger
archived in the RFC's Alternatives/Deferred; the interim guidance-only
non-shell restraint is priced in Consequences.

Exiting plan is now a pure removal (the exit tool + section), which the
delta encoding CAN express: the re-recorded plan-mode fixture pins one
plan-shaped initial header snapshot plus one header-delta instead of
two snapshots.
2026-07-12 23:20:36 +08:00
Tianyi Cui
2ca806a1ec docs: refresh scoped lifecycle catalogs 2026-07-12 23:16:30 +08:00
Tianyi Cui
738054562d fix: complete scoped lifecycle simplification 2026-07-12 23:15:07 +08:00
kingwl
99650a201b feat(mode): the access cap — plan mode composes with the sandbox instead of banning bash
A ModeDefinition may declare access: the widest sandbox access shell
commands run under while the mode holds, on the SANDBOX_MODES ladder.
The bash seam gains the resolution point to hang it on: BashExecutor.
resolveMode(session) folds override ?? default and dispatches the new
bash/resolve-mode waterfall; dsh-tool-bash consults it at both the
stamping site and the escalation baseline; dsh-mode's clamp listener
takes the ladder minimum per call. Two independent log folds compose at
read time — the mode never writes the sandbox knob, so the two switch
in any order and the knob re-emerges intact on exit.

The built-in plan definition ships access: read-only with the bash trio
allowlisted CONDITIONALLY: both policy layers admit bash/bash_output/
bash_kill only while a confining executor is mounted (an unconfinable
shell cannot honor the cap), and a bash call carrying sandbox_permissions
under a cap is denied at the gate — no widening mid-mode; the widened
step belongs in the plan.

examples/plan-acp-agent swaps bash-local for sandbox-local +
bash-sandbox (workspace-write default, clamped read-only inside plan)
plus the approval seam; the re-recorded plan-mode arc runs a real cat
inside plan under the clamped sandbox, and modes-advertise now pins the
sandbox-mode and approval config options. RFC amended to the landed
shape (access cap section, orthogonality FAQ, deferred item resolved
into effects self-declaration).
2026-07-12 22:51:09 +08:00
Tianyi Cui
ed5304fb6d docs(rfc): align scoped runtime contracts 2026-07-12 22:49:46 +08:00
Tianyi Cui
bb3f6bd736 refactor(subagent): unify async readiness and cancellation 2026-07-12 22:41:59 +08:00
Tianyi Cui
02ca71db57 refactor(core): simplify tools prompts and trusted services 2026-07-12 22:39:01 +08:00
Tianyi Cui
28e04ff4fb refactor(core): simplify scoped agent lifecycles 2026-07-12 22:36:04 +08:00
kingwl
88db403d9f Merge branch 'worktree-session-modes-rfc' (master: sandbox stack #169, skills #109, prompt snapshots #254)
The stack rebases onto a moved master through its base branch. Beyond
mechanical unions (both branches' demo scripts, example rows, service
roles, tool lists, acp deps, doc budgets — each side fit alone, the
union needs the higher ceilings), three semantic reconciliations:

- The ACP bridge now carries BOTH per-session surfaces: the sandbox
  stack's config options + approval answerer and this branch's session
  modes; session/new and session/load advertise modes AND configOptions
  side by side.
- The feature matrix supersedes the sandbox stance per the RFC's
  second-lander rule: session/set_mode and current_mode_update flip to
  shipped-by-dsh-mode, config-option rows stay as #169 wrote them, and
  §6 records both landed features under the picker-to-modes /
  knobs-to-config-options division.
- The snapshot pin grammar (#254: one header snapshot + declared deltas
  + a Markdown prompt golden) gains a symmetric declaration for what a
  delta cannot express: expectedHeaderSnapshots — a plan-mode flip
  resorts the canonical tool list, so its widening lands as a second
  full snapshot, now its own Markdown section. The pin-less-class and
  model-turn-only-pin amendments carry over; new fixtures cover the
  extended writer paths, and the plan-acp-agent scenarios re-recorded
  under the merged composition (the app now bundles the skill tool)
  with the suite's refresh mode wired through.
2026-07-12 20:08:09 +08:00
kingwl
2fc1e0bad5 Merge remote-tracking branch 'origin/master' into worktree-session-modes-rfc 2026-07-12 19:42:29 +08:00
Tianyi Cui
e8fed4fb66 fix(session): contain post-commit observers 2026-07-12 18:57:42 +08:00
Tianyi Cui
50873b8bd0 docs(rfc): remove redundant lifecycle diagram 2026-07-12 18:10:04 +08:00
Tianyi Cui
11a074b664 docs(rfc): add agent-scope diagrams 2026-07-12 17:21:15 +08:00
Tianyi Cui
a34801df4b fix(agent-loop): own queued message input 2026-07-12 16:54:37 +08:00
Tianyi Cui
7e3d46a3ce docs(scope): split contract from runtime design 2026-07-12 13:25:04 +08:00
Tianyi Cui
c6d012109f docs(scope): restructure RFC top-down 2026-07-12 12:22:19 +08:00
Tianyi Cui
cb03c8c284 fix(scope): align trust and input boundaries
Rewrite the agent-scope RFC with executable examples and an explicit security non-goal. Harden subagent scalar and depth validation, and pin live tool-filter semantics across code, tests, and generated docs.
2026-07-12 11:17:57 +08:00
Tianyi Cui
d427478c44 fix(subagent): validate direct depth boundaries 2026-07-12 10:48:20 +08:00
Tianyi Cui
875c3d62d8 fix(workflow): reap children after settled dispose 2026-07-12 10:39:55 +08:00
Tianyi Cui
48067c3a7a fix(subagent): validate depth config at load 2026-07-12 10:33:29 +08:00
Tianyi Cui
9fc2260bb6 fix(workflow): harden terminal cleanup races
Queue worker results before settlement cleanup, claim terminal and death boundaries before provider callbacks, and close late-message admission.

Make child cancellation and disposal reentrancy-safe across the workflow bridge and generic subagent wrapper, with adversarial regression coverage and RFC documentation.
2026-07-12 10:17:31 +08:00
Tianyi Cui
c5b1a7941f fix(scope): harden lifecycle ownership foundation
Make Cordis construction and teardown ownership reentrancy-safe, then carry caller and provider ownership through reservation, setup, publication, quiescence, and sentinel retirement.

Stabilize registry carriers and factory/workflow boundaries, add adversarial lifecycle regressions, and align the rewritten RFC plus generated contracts with the enforced behavior.
2026-07-12 08:57:05 +08:00
Tianyi Cui
197f7237d2 fix(workflow): bootstrap source worker transforms 2026-07-12 06:19:53 +08:00
Tianyi Cui
a9cb70d896 fix(scope): harden final ownership boundaries 2026-07-12 05:13:17 +08:00
Tianyi Cui
36b8370027 fix(scope): close remaining ownership boundaries 2026-07-12 03:51:55 +08:00
Tianyi Cui
3dca90261c Merge remote-tracking branch 'origin/master' into codex/pr224-rfc-rewrite
# Conflicts:
#	examples/coding-agent/tests/code-mode-keyless-smoke.e2e.ts
#	examples/coding-agent/tests/keyless-smoke.e2e.ts
#	examples/cordis-agent/tests/keyless-smoke.e2e.ts
#	examples/echo-agent/tests/echo.e2e.ts
2026-07-12 01:39:29 +08:00
Tianyi Cui
4cda9dd03d Merge pull request #254 from deepseek-harness/codex/system-prompt-snapshots
test(acp): snapshot system prompts as Markdown
2026-07-12 01:36:31 +08:00
Tianyi Cui
35715423f8 fix(workflow): honor subagent readiness boundary 2026-07-12 01:35:09 +08:00
Tianyi Cui
c632b693c6 test(e2e): tolerate bounded loader contention 2026-07-12 00:50:50 +08:00
Tianyi Cui
cb05300ba7 docs: remove duplicate scoped-event explanation 2026-07-12 00:38:38 +08:00
Tianyi Cui
3529b3c166 fix(scope): close final ownership races
Drain idle injection flushes before agent teardown, snapshot approval and subagent provider inputs, and gate subagent lifecycle events on real child readiness. Align the RFCs and generated contracts with the hardened behavior.
2026-07-12 00:31:46 +08:00
Tianyi Cui
382c243e78 test(e2e): widen keyless boot-smoke kill budgets to 30s
The keyless Loader-path smokes killed the child 10s after spawn, but a
loaded CI e2e runner routinely needs longer just to boot the unbuilt tsx
tree: on this branch's run the coding-agent smoke burned both retries and
failed at 30s wall-clock, and the sibling smokes passed only on retry x2
(master's latest run shows the same near-misses). The budget guards
against a HANG, not slowness — raise kill to 30s and the vitest test
timeout to 45s so a slow boot no longer masquerades as one.
2026-07-11 23:56:16 +08:00
Tianyi Cui
bac97a5b2e test(acp-snapshot): cover the delta-insert trailing-newline branch
formatSystemPromptSnapshot's insert-join ternary had its already-newline-
terminated arm unexercised (a delta whose insert ends in a blank line),
failing the per-file 100% branch gate on suite.ts (99.24%).
2026-07-11 23:55:56 +08:00
Tianyi Cui
e6c7522a75 fix(agent-loop): nest config resume teardown 2026-07-11 23:45:01 +08:00
Tianyi Cui
06e42f439c docs: regenerate merged catalogs 2026-07-11 23:42:52 +08:00
Tianyi Cui
cf255eebb1 fix(scope): harden merged tool and skill boundaries 2026-07-11 23:41:37 +08:00
Tianyi Cui
172005e0e6 fix(scope): align merged agent event contracts 2026-07-11 23:22:57 +08:00
Tianyi Cui
56e34619d7 Merge remote-tracking branch 'origin/master' into codex/pr224-rfc-rewrite
# Conflicts:
#	docs/architecture.md
#	docs/capability-seams.md
#	docs/config-catalog.md
#	docs/cookbook/extension-cookbook.md
#	docs/cordis-catalog/events.md
#	docs/cordis-catalog/services.md
#	docs/event-producer-consumer.md
#	docs/module-graph.md
#	docs/rfc/implemented/feature/2026-06-30-interception-seams.md
#	docs/rfc/proposed/feature/2026-06-14-acp-agent-client-protocol.md
#	docs/tool-execution-pipeline.md
#	packages/cordis/tool-cordis/src/api-catalog.ts
#	packages/core/agent-core/README.md
#	packages/core/agent-loop/README.md
#	packages/core/tools/README.md
#	packages/core/tools/src/index.ts
#	packages/core/tools/tests/tools.spec.ts
#	packages/core/tools/tsconfig.json
#	packages/ui/acp/src/index.ts
#	scripts/doc-budgets.manifest.json
#	scripts/gen-cordis-catalog.ts
#	scripts/gen-doc-graphs.ts
2026-07-11 23:14:09 +08:00
Tianyi Cui
36c94fbe3e docs: align the agent-scope contracts 2026-07-11 22:55:40 +08:00
Tianyi Cui
3263dab822 fix(core): enforce agent-scoped ownership boundaries 2026-07-11 22:55:26 +08:00
Tianyi Cui
da9cef381c Merge remote-tracking branch 'origin/master' into codex/system-prompt-snapshots
# Conflicts:
#	examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl
#	examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl
#	examples/acp-agent/tests/snapshots/text-turn/session.jsonl
#	packages/support/acp-snapshot/src/suite.ts
2026-07-11 22:48:14 +08:00
Tianyi Cui
7331720c8a Merge pull request #109 from deepseek-harness/codex/skill-system
Add skill registry and local provider
2026-07-11 22:35:05 +08:00
Tianyi Cui
169f86966a fix(review): lint prompt Markdown paragraphs 2026-07-11 22:33:33 +08:00