Commit Graph

56 Commits (4329d94de3a71803db177ee50af68eb75d7ea06b)

Author SHA1 Message Date
Peter Steinberger 0b159d7250
fix(test): restore support shard boundaries 7 months ago
Vincent Koc f7dc5f930a fix(plugin-sdk): add runtime boundary entrypoints 8 months ago
Peter Steinberger 087eb621ff
build: fix plugin sdk boundary exports 8 months ago
Peter Steinberger ad0c4309e6
refactor: dedupe shared trim readers 8 months ago
Peter Steinberger 9a66b9cd54 Tests: fix package boundary and runtime drift 8 months ago
Peter Steinberger 7cf72f7bc8 Tests: skip bedrock auth probe in embeddings spec 8 months ago
Peter Steinberger 67c4733267
build: align plugin sdk package boundaries 8 months ago
Peter Steinberger 8c7dd66a7b
refactor: dedupe string readers 8 months ago
Peter Steinberger 2f115bc645
refactor: dedupe reader helpers 8 months ago
Peter Steinberger d03985415d
refactor: dedupe trimmed string readers 8 months ago
Peter Steinberger b7be963501
refactor: dedupe record guards 8 months ago
Peter Steinberger 69f4022950
refactor: dedupe browser and memory host error helpers 8 months ago
Peter Steinberger fbebf6147c Tests: fix boundary drift and shell preflight regressions 8 months ago
Peter Steinberger 80c8567f9d
fix: resolve merge conflicts and preserve runtime test fixes 8 months ago
Peter Steinberger ed64ce3983
build: type plugin sdk exports for xai 8 months ago
Harold Hunt 0bd0097557
refactor: add xai plugin-sdk boundary canary (#61548)
* docs: plan real plugin-sdk workspace rollout

* build: add xai plugin-sdk boundary canary

* build: generate plugin-sdk package types

* build: hide plugin-sdk core export

* build: alias scoped plugin-sdk runtime imports

* build: repair plugin-sdk boundary drift

* fix(plugins): remove duplicated plugin-sdk entrypoints

* test(plugins): make tsc boundary canary portable

---------

Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
8 months ago
wirjo 699b2320a8
feat(memory): add Bedrock embedding provider for memory search (#61547)
* feat(memory): add Bedrock embedding provider for memory search

Add Amazon Bedrock as a native embedding provider for memory search.
Supports Titan Embed Text v1/v2 and Cohere Embed models via AWS SDK.

- New embeddings-bedrock.ts: BedrockRuntimeClient + InvokeModel
- Auth via AWS default credential chain (same as Bedrock inference)
- Auto-selected in 'auto' mode when AWS credentials are detected
- Titan V2: configurable dimensions (256/512/1024), normalization
- Cohere: native batch support with search_query/search_document types
- 16 new tests covering all model types, auth detection, edge cases

Closes #26289

* fix(memory): harden bedrock embedding selection

---------

Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
8 months ago
Dave Morin 2ed2dbba00
Memory: move dreaming trail to dreams.md (#61537)
* Memory: move dreaming trail to dreams.md

* docs(changelog): add dreams.md entry

---------

Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
8 months ago
Peter Steinberger 2aafa8fb7d
refactor: remove ollama sdk facades 8 months ago
Vincent Koc 2a999bf9c9 refactor(memory): invert memory host sdk dependency 8 months ago
scoootscooob 9860db5cea
fix(memory): allow Gemini multimodal fallback before registry hydration (#61085)
* fix(memory): allow Gemini multimodal fallback

* docs(memory): clarify multimodal fallback
8 months ago
Peter Steinberger bc23db501b
test: trim more core importOriginal usage 8 months ago
Shakker e9e7033ea1 test: trim embeddings provider import cost 8 months ago
Peter Steinberger 847faa3d04
test: trim extension test import churn 8 months ago
ImLukeF 78b48735fa
test: restore fetch stubs in embedding suites 8 months ago
ImLukeF 101c31f5e1
test: harden ci-sensitive unit suites 8 months ago
Peter Steinberger 759d37635d
Revert "refactor: move tasks behind plugin-sdk seam"
This reverts commit da6e9bb76f.
8 months ago
Peter Steinberger a1e2d2bf42
test: repair stale task and image mocks 8 months ago
Peter Steinberger da6e9bb76f
refactor: move tasks behind plugin-sdk seam 8 months ago
Vincent Koc ab4ddff7f1
feat(memory): add per-agent QMD extra collections for cross-agent session search (#58211)
* feat(memory): add per-agent qmd extra collections

* test(config): cover qmd extra collections schema outputs

* docs(config): refresh qmd extra collections baseline

* docs(config): regenerate qmd extra collections baselines

* docs(config): clarify qmd extra collection naming
8 months ago
Vincent Koc 075645f5cb
fix(memory): use explicit qmd snippet line metadata (#58181)
* fix(memory): preserve qmd snippet line metadata

* Memory/QMD: preserve snippet span with partial line metadata
8 months ago
Altay 910134b702 fix(memory): stabilize qmd collection scoping 8 months ago
Vincent Koc b7de04f23f
fix(memory): preserve shared qmd collection names (#57628)
* fix(memory): preserve shared qmd collection names

* fix(memory): canonicalize qmd path containment
8 months ago
Vincent Koc 8623c28f1d
fix(memory): warn when qmd binary is missing (#57467)
* fix(memory): warn when qmd binary is missing

* fix(memory): avoid probing cached qmd managers

* docs(memory): clarify qmd doctor probe behavior

* fix(memory): probe qmd from agent workspace
8 months ago
Vincent Koc b7d59f7831
fix(memory): export archived qmd session transcripts (#57446)
* fix(memory): export archived qmd session transcripts

* test(memory): separate qmd session listing describe
8 months ago
Gustavo Madeira Santana 313fdf5adf
Memory Host: make backend-config path tests portable 8 months ago
Vincent Koc da35718cb2
fix(memory): add qmd mcporter search tool override (#57363)
* fix(memory): add qmd mcporter search tool override

* fix(memory): tighten qmd search tool override guards

* chore(config): drop generated docs baselines from qmd pr

* fix(memory): keep explicit qmd query override on v2 args

* docs(changelog): normalize qmd search tool attribution

* fix(memory): reuse v1 qmd tool after query fallback
8 months ago
Amine Harch el korane 219d4f03bd
fix: wire memorySearch.extraPaths to QMD indexing (#57315)
* fix: wire memorySearch.extraPaths to QMD indexing

The 'agents.defaults.memorySearch.extraPaths' config field was documented
to add extra directories to the memory index, but the paths were never
actually passed to the QMD backend. Only 'memory.qmd.paths' worked.

This fix reads extraPaths from the memorySearch config and maps them
to QMD custom path collections, so users can simply configure:

  memorySearch:
    extraPaths:
      - odd-vault
      - /Users/odd/workspace
      - /Users/odd/docs

And have those directories indexed alongside the default memory files.

Closes #57302

* fix: handle per-agent memorySearch.extraPaths overrides + add tests

- Read per-agent overrides from agents.list[].memorySearch.extraPaths
- Agent-specific overrides take priority over defaults
- Falls back to defaults when agent has no overrides
- Added 3 test cases for the feature

* fix: merge defaults + agent overrides instead of replacing

* fix: remove any types from tests, fix merge behavior assertion

* fix(memory): merge qmd extra path collections

* fix(memory): normalize qmd extra path resolution

* fix(memory): type qmd extra path merge

---------

Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
8 months ago
George Zhang e133924047
[codex] harden clawhub plugin publishing and install (#56870)
* fix: harden clawhub plugin publishing and install

* fix(process): preserve windows shim exit success
8 months ago
Peter Steinberger 8e0ab35b0e
refactor(plugins): decouple bundled plugin runtime loading 8 months ago
Tak Hoffman 3ce48aff66
Memory: add configurable FTS5 tokenizer for CJK text support (openclaw#56707)
Verified:
- pnpm build
- pnpm check
- pnpm test -- extensions/memory-core/src/memory/manager-search.test.ts packages/memory-host-sdk/src/host/query-expansion.test.ts
- pnpm test -- extensions/memory-core/src/memory/index.test.ts -t "reindexes when extraPaths change"
- pnpm test -- src/config/schema.base.generated.test.ts
- pnpm test -- src/media-understanding/image.test.ts
- pnpm test

Co-authored-by: Mitsuyuki Osabe <24588751+carrotRakko@users.noreply.github.com>
8 months ago
AaronLuo00 f8547fcae4 fix: guard fine-split against breaking UTF-16 surrogate pairs
When re-splitting CJK-heavy segments at chunking.tokens, check whether the
slice boundary falls on a high surrogate (0xD800–0xDBFF) and if so extend
by one code unit to keep the pair intact.  Prevents producing broken
surrogate halves for CJK Extension B+ characters (U+20000+).

Add test verifying no lone surrogates appear when splitting lines of
surrogate-pair characters with an odd token budget.

Addresses third-round Codex P2 review comment.
8 months ago
AaronLuo00 3b95aa8804 fix: address second-round review — Latin backward compat and emoji consistency
- Two-pass line splitting: first slice at maxChars (unchanged for Latin),
  then re-split only CJK-heavy segments at chunking.tokens. This preserves
  the original ~800-char segments for ASCII lines while keeping CJK chunks
  within the token budget.

- Narrow surrogate-pair adjustment to CJK Extension B+ range (D840–D87E)
  only, so emoji surrogate pairs are not affected. Mixed CJK+emoji text
  is now handled consistently regardless of composition.

- Add tests: emoji handling (2), Latin backward-compat long-line (1).

Addresses Codex P1 (oversized CJK segments) and P2s (Latin over-splitting,
emoji surrogate inconsistency).
8 months ago
AaronLuo00 a5147d4d88 fix: address bot review — surrogate-pair counting and CJK line splitting
- Use code-point length instead of UTF-16 length in estimateStringChars()
  so that CJK Extension B+ surrogate pairs (U+20000+) are counted as 1
  character, not 2 (fixes ~25% overestimate for rare characters).

- Change long-line split step from maxChars to chunking.tokens so that
  CJK lines are sliced into token-budget-sized segments instead of
  char-budget-sized segments that produce ~4x oversized chunks.

- Add tests for both fixes: surrogate-pair handling and long CJK line
  splitting.

Addresses review feedback from Greptile and Codex bots.
8 months ago
AaronLuo00 971ecabe80 fix(memory): account for CJK characters in QMD memory chunking
The QMD memory system uses a fixed 4:1 chars-to-tokens ratio for chunk
sizing, which severely underestimates CJK (Chinese/Japanese/Korean) text
where each character is roughly 1 token. This causes oversized chunks for
CJK users, degrading vector search quality and wasting context window space.

Changes:
- Add shared src/utils/cjk-chars.ts module with CJK-aware character
  counting (estimateStringChars) and token estimation helpers
- Update chunkMarkdown() in src/memory/internal.ts to use weighted
  character lengths for chunk boundary decisions and overlap calculation
- Replace hardcoded estimateTokensFromChars in the context report
  command with the shared utility
- Add 13 unit tests for the CJK estimation module and 5 new tests for
  CJK-aware memory chunking behavior

Backward compatible: pure ASCII/Latin text behavior is unchanged.

Closes #39965
Related: #40216
8 months ago
Peter Steinberger 4c27c90fc2 refactor: finish moving provider runtime into extensions 8 months ago
Peter Steinberger 64bf80d4d5 refactor: move provider runtime into extensions 8 months ago
Peter Steinberger eebce9e9c7
refactor: move memory host into sdk package 8 months ago
Peter Steinberger bd6c7969ea
refactor: extract memory host sdk package 8 months ago
Peter Steinberger 7695b4842b chore: bump version to 2026.2.12 9 months ago