qwen-code/docs/users
Dragon 18cb393e4b
feat(core): preload deferred tools within a context-window threshold (#7922)
* feat(core): preload deferred tools within a context-window threshold

Adds tools.toolSearch.threshold (default 10, percent of the context
window). At session start, when the combined estimated schema footprint
of every deferred tool - bundled built-ins and MCP alike - fits within
the budget, all are revealed upfront so the declaration list stays
stable for the whole session and prefix KV caches survive; otherwise
everything stays deferred. Set 0 to always defer. Mirrors Claude Code's
ENABLE_TOOL_SEARCH=auto threshold mode, extended to bundled deferred
tools because here every reveal rewrites the declaration list and busts
the prompt-cache prefix.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(core): log deferred-tool preload budget decision

Emit debugLogger diagnostics in preloadDeferredToolsWithinBudget covering
the computed budget, estimated token footprint, candidate count, and which
branch of the all-or-nothing gate was taken (no candidates, over budget, or
preloaded). Lets an operator diagnosing session-startup cost tell from debug
logs whether the deferred set fit the budget or was left behind ToolSearch,
without adding temporary instrumentation. No behavior change.

* fix(tools): bound toolSearch.threshold to 0-100%

The threshold setting is a percentage of the context window but had no
upper bound, so a value like 200 (a typo or misreading of the "(%)"
label) made the preload budget exceed the whole window and
unconditionally preloaded every deferred tool — the opposite of the
prefix-stability the threshold buys.

- Add minimum:0/maximum:100 to the setting schema (jsonSchemaOverride,
  like autoCompactThreshold) and regenerate the VS Code settings schema.
- Add a symmetric runtime upper guard next to the existing
  'thresholdPercent <= 0' lower guard in client.ts, clamping to 100% so a
  hand-edited settings file cannot slip a larger budget past validation.

Adds a client test asserting a 200% threshold clamps to a full-context
budget.

* test(core): cover configured preload budget

* fix(tool-search): harden preload threshold

* test(tool-search): cover preload exclusions

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 23:01:01 +00:00
..
configuration feat(core): preload deferred tools within a context-window threshold (#7922) 2026-07-29 23:01:01 +00:00
extension fix(core): Align MCP OAuth guidance and docs (#5589) 2026-06-24 07:09:53 +08:00
features fix(channels): reject unusable GitHub self-allowlists (#8055) 2026-07-29 15:48:56 +00:00
ide-integration fix(cli): replace all emoji with Unicode text symbols in TUI rendering (#5999) 2026-06-30 15:18:01 +00:00
reference fix(cli): default to virtualized terminal history (#5738) 2026-07-28 12:51:55 +00:00
support fix(cli): default to virtualized terminal history (#5738) 2026-07-28 12:51:55 +00:00
_meta.ts feat(daemon): merge daemon-mode feature batch into main (#4490) 2026-06-12 00:34:49 +08:00
common-workflow.md docs: fix stale defaults, CLI syntax, and tool naming drift (#5158) 2026-06-15 20:06:34 +08:00
integration-github-action.md feat(web-shell): show subagent sessions in detail panel (#7380) 2026-07-22 02:31:08 +00:00
integration-jetbrains.md docs(integration): use CDN URLs for images and fix formatting 2026-03-16 14:12:48 +08:00
integration-vscode.md fix(core): honor NO_PROXY for model requests (#6640) 2026-07-10 10:41:04 +00:00
integration-zed.md docs(integration): use CDN URLs for images and fix formatting 2026-03-16 14:12:48 +08:00
overview.md docs: fix config/command/auth drift and surface the model-providers page (#5735) 2026-06-24 06:06:01 +08:00
quickstart.md docs: fix config/command/auth drift and surface the model-providers page (#5735) 2026-06-24 06:06:01 +08:00
qwen-serve-deploy-local.md docs(serve): Close multi-workspace hardening gaps (#7019) 2026-07-16 17:33:53 +00:00
qwen-serve.md fix(serve): allow bounded reads of large text files (#7947) 2026-07-29 07:49:51 +00:00