roboomp 38454d3321 fix(agent): excluded orchestration tokens from context sizing ...

calculateContextTokens returned usage.totalTokens which, with the new
Usage.orchestration sidecar, folds provider-side orchestration back into the
context size used by auto-compaction/context promotion thresholds. Subtract
the orchestration sidecar so context sizing stays conversation-only while
cost and totalTokens keep the orchestration spend visible.

Refs #4469

2026-07-03 16:53:11 +00:00

..

utils

feat(coding-agent/tools): added explicit ArkType schema descriptions to agent tools

2026-06-18 00:59:55 +02:00

agent-loop.test.ts

Merge remote-tracking branch 'origin/farm/6f26b11c/distinguish-synthetic-tool-failure'

2026-07-02 23:42:38 +02:00

agent-side-request-context.test.ts

feat: migrated core rendering and context transformation to async

2026-06-21 00:32:13 +02:00

agent.test.ts

Merge PR #3060: feat(provider): add GitLab Duo Agent provider (@jiwangyihao)

2026-06-27 01:39:31 +02:00

append-only-context.test.ts

fix(agent): include provider payloads in append-only digests

2026-06-24 23:11:12 +00:00

branch-summarization.test.ts

fix(agent): budget branch tool results after truncation

2026-07-01 21:53:15 +02:00

compaction-error-status.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

compaction-file-ops.test.ts

feat(agent): prevented internal URLs from appearing in file summaries

2026-06-18 19:38:31 +02:00

compaction-reserve-provenance.test.ts

fix(agent): honored explicit compaction reserve equal to default

2026-07-02 00:32:48 +02:00

compaction-telemetry.test.ts

feat: added build-time compatibility in ModelSpec/buildModel pipeline

2026-06-10 06:20:51 +02:00

compaction-thinking-level.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

context-tokens-orchestration.test.ts

fix(agent): excluded orchestration tokens from context sizing

2026-07-03 16:53:11 +00:00

handoff.test.ts

feat: optimized handoff generation and harden request safety

2026-06-22 20:05:40 +02:00

helpers.ts

test(agent): replaced agent stream mocks with createMockModel responses

2026-05-15 14:46:54 +02:00

normalize-tools-prune.test.ts

fix(agent): scoped intent injection inside union variants

2026-06-27 11:31:11 +00:00

otel.test.ts

fix(agent): handle string systemPrompt in telemetry content capture

2026-06-04 18:51:31 +00:00

prompt-tools-loop.test.ts

docs(ai/dialect): clarified escaping and halting rules for tool dialects

2026-06-26 22:35:20 +02:00

proxy-stream-disconnect.test.ts

feat(agent): resolved partial json leakage by enforcing streaming cleanup

2026-06-27 12:26:11 +02:00

proxy-toolcall-partial-json.test.ts

feat: implemented symbol-based streaming state for isolated metadata

2026-06-27 12:07:08 +02:00

remote-compaction.test.ts

test(agent): strengthened consistency and persistence logic in integration tests

2026-06-28 09:52:44 +02:00

run-summary.test.ts

feat(agent): updated steering tests and clarified task batching guidance

2026-06-12 03:31:40 +02:00

serialize-conversation.test.ts

refactor: renamed grammar module to dialect with unified transcript rendering

2026-06-15 13:58:12 +02:00

shake.test.ts

feat(agent): added useless-result tagging and compaction dropping of tool outputs

2026-06-12 16:26:38 +02:00

snapcompact-frames.test.ts

refactor(snapcompact): dropped the snapcompact qualifier from every export

2026-06-12 03:27:50 +02:00

soft-tool-requirement.test.ts

feat: implemented soft tool requirement and escalation system

2026-06-19 07:51:27 +02:00

supersede-prune.test.ts

perf(compaction): kept per-turn tool-result pruning inside the warm prompt-cache prefix

2026-06-19 06:57:53 +02:00

tool-protection.test.ts

fix(agent/compaction): skipped truncating tiny tool outputs during pruning

2026-06-15 12:45:21 +02:00

yield.test.ts

test(agent): migrated yield tests to use YieldGate for stability

2026-06-24 15:30:29 +02:00