feat(llm): send an explicit reasoning_effort for Kimi models that support it (GRAPHIFY_KIMI_EFFORT) - #2982

Closed

Azeem1985 wants to merge 21 commits into

Graphify-Labs:v8from

TelB-io:feat/kimi-reasoning-effort

Commits on Aug 23, 2026

feat(llm): openai-cli backend — semantic extraction through the locally authenticated Codex CLI
Azeem
committed
2d6130a
feat(llm): kimi reasoning_effort — send an explicit effort (GRAPHIFY_KIMI_EFFORT, default max) for models that support it
Azeem
committed
417ec1e
feat(cli): credential gate knows openai-cli — a present codex CLI is the credential (OAuth), mirroring claude-cli
Azeem
committed
c771b67
fix(openai-cli): disable each configured MCP server per call — a blanket mcp_servers={} does not work
Azeem
committed
d7c8eb8

Commits on Aug 24, 2026

feat: extract holds the per-repo rebuild lock for its whole pipeline

Azeem
and
claude
committed
f7c5899
feat: extract --fallback-backend retries a zero-success semantic pass on a second backend

Azeem
and
claude
committed
627a42b
feat: graphify watch --semantic runs LLM extraction automatically on doc changes

Azeem
and
claude
committed
83e12ef
feat: batched UNWIND push for neo4j/falkordb export (--batch-size)

Azeem
and
claude
committed
65a187f
test: a partially-succeeded primary pass must not fire the fallback

Azeem
and
claude
committed
6269f91
Merge pull request #1 from TelB-io/feat/extract-lock
Azeem1985
authored
2373895
Merge remote-tracking branch 'fork/v8' into feat/fallback-backend
Azeem
committed
76b928a
Merge pull request #2 from TelB-io/feat/fallback-backend
Azeem1985
authored
dedb7ff
Merge remote-tracking branch 'fork/v8' into feat/watch-semantic
Azeem
committed
fe5682d
Merge pull request #3 from TelB-io/feat/watch-semantic
Azeem1985
authored
7f13af8
Merge remote-tracking branch 'fork/v8' into feat/batched-db-push
Azeem
committed
ddefdf1
Merge pull request #4 from TelB-io/feat/batched-db-push
Azeem1985
authored
7538bd8
Merge branch 'v8' into feat/openai-cli-backend
Azeem
committed
243280a
test(openai-cli): cover the happy parse, failure raises with vendor detail, and turn-usage accounting; changelog

Azeem
and
claude
committed
1d0a933
Merge pull request #5 from TelB-io/feat/openai-cli-backend
Azeem1985
authored
3788ca7
Merge branch 'v8' into feat/kimi-reasoning-effort
Azeem
committed
89b1c3a
test(kimi): reasoning_effort assembles from GRAPHIFY_KIMI_EFFORT with default max; changelog

Azeem
and
claude
committed
f4145cf

Navigation Menu

Uh oh!

Uh oh!

feat(llm): send an explicit reasoning_effort for Kimi models that support it (GRAPHIFY_KIMI_EFFORT) - #2982

feat(llm): send an explicit reasoning_effort for Kimi models that support it (GRAPHIFY_KIMI_EFFORT)#2982
Azeem1985 wants to merge 21 commits into
Graphify-Labs:v8from
TelB-io:feat/kimi-reasoning-effort

Commits on Aug 23, 2026

feat(llm): openai-cli backend — semantic extraction through the locally authenticated Codex CLI

feat(llm): kimi reasoning_effort — send an explicit effort (GRAPHIFY_KIMI_EFFORT, default max) for models that support it

feat(cli): credential gate knows openai-cli — a present codex CLI is the credential (OAuth), mirroring claude-cli

fix(openai-cli): disable each configured MCP server per call — a blanket mcp_servers={} does not work

Commits on Aug 24, 2026

feat: extract holds the per-repo rebuild lock for its whole pipeline

feat: extract --fallback-backend retries a zero-success semantic pass on a second backend

feat: graphify watch --semantic runs LLM extraction automatically on doc changes

feat: batched UNWIND push for neo4j/falkordb export (--batch-size)

test: a partially-succeeded primary pass must not fire the fallback

Merge pull request #1 from TelB-io/feat/extract-lock

Merge remote-tracking branch 'fork/v8' into feat/fallback-backend

Merge pull request #2 from TelB-io/feat/fallback-backend

Merge remote-tracking branch 'fork/v8' into feat/watch-semantic

Merge pull request #3 from TelB-io/feat/watch-semantic

Merge remote-tracking branch 'fork/v8' into feat/batched-db-push

Merge pull request #4 from TelB-io/feat/batched-db-push

Merge branch 'v8' into feat/openai-cli-backend

test(openai-cli): cover the happy parse, failure raises with vendor detail, and turn-usage accounting; changelog

Merge pull request #5 from TelB-io/feat/openai-cli-backend

Merge branch 'v8' into feat/kimi-reasoning-effort

test(kimi): reasoning_effort assembles from GRAPHIFY_KIMI_EFFORT with default max; changelog