Skip to content

docs: align SDK and provider guidance with installed dependencies - #380

Merged
developerz-ai[bot] merged 1 commit into
mainfrom
dz/align-sdk-and-provider-house-guidance-with--28d04c9e
Nov 3, 2026
Merged

developerz-ai[bot] merged 1 commit into
mainfrom
dz/align-sdk-and-provider-house-guidance-with--28d04c9e

Conversation

@developerz-ai

@developerz-ai developerz-ai Bot commented Nov 3, 2026 •

Copy link
Copy Markdown
Contributor

What

The root CLAUDE.md stack table still named ai ^6.0.182 and @ai-sdk/mcp ^2.0.87, and its
provider note ordered the hoisted @ai-sdk/provider copy to stay on the v3 line. The manifests
and bun.lock are right: ai v7 and @ai-sdk/mcp v2 both resolve provider v4 (4.0.35) and
@openrouter/ai-sdk-provider v3 declares no provider dependency, so no v3 has to be defended.

Restate the AI SDK row for v7 (ToolLoopAgent; experimental_Agent survives only as an alias
export), rewrite the provider-hoisting paragraph around the current relationship, and fix the
ai-claude-compat README's stale "Peer dep: ai (AI SDK v6)" line — ai is a direct dependency
at ^7.0.159. Prose only: no dependency version moves.

Acceptance

  • Root CLAUDE.md's ## Stack table shows ai ^7.0.159 and @ai-sdk/mcp ^2.0.88, matching packages/aitm/package.json lines 46-49.
  • Root CLAUDE.md's provider paragraph states that ai v7 and @ai-sdk/mcp v2 both resolve @ai-sdk/provider v4 (4.0.35) and that @openrouter/ai-sdk-provider v3 declares no provider dependency; no line in the file still directs the reader to hold the hoisted copy at provider v3.
  • The AI SDK row no longer describes the SDK as v6 and no longer claims experimental_Agent is gone — it states the v7 naming: ToolLoopAgent, with experimental_Agent only an alias export of it.
  • packages/ai-claude-compat/README.md states ai as this package's direct dependency ^7.0.159 (AI SDK v7) instead of 'Peer dep: ai (AI SDK v6)'.
  • The diff touches only CLAUDE.md and packages/ai-claude-compat/README.md — no package.json, bun.lock, or vendored file is modified, so no dependency version is changed to match stale prose.
  • CLAUDE.md's 'Releases and bun.lock' measured-on-chore(deps): bun.lock's workspace versions are eight releases stale, and regenerating it is a hidden AI SDK upgrade #327 version list (ai 6.0.182 → 6.0.240, @ai-sdk/provider 3.0.10 → 3.0.14) is left verbatim — it is a historical record, not current guidance.
  • bun run lint (biome check .) passes on the tree with the two edited markdown files.

Changes

  • CLAUDE.md (edit) — Rewrite two spots of the house guidance: (1) the ## Stack table's AI SDK row — ai ^6.0.182 → ^7.0.159, @ai-sdk/mcp ^2.0.87 → ^2.0.88, and replace the stale '(v6; the AI SDK 5 name experimental_Agent is gone)' with the v7 naming (ToolLoopAgent; experimental_Agent exists in v7 only as an alias export of it). (2) the 'Provider & AI SDK' bullet that orders the reader to keep the hoisted @ai-sdk/provider on the v3 line — replace it with the current relationship (ai v7 and @ai-sdk/mcp v2 both resolve provider v4 4.0.35; @openrouter/ai-sdk-provider v3 declares no provide [trimmed]
  • packages/ai-claude-compat/README.md (edit) — Fix the '## Runtime' sentence 'Peer dep: ai (AI SDK v6)': ai is a DIRECT dependency of this package (ai: ^7.0.159 → AI SDK v7) — packages/ai-claude-compat/package.json declares no peerDependencies block at all. Rewrite the sentence to say the direct dependency and AI SDK v7; nothing else in the file changes.
  • packages/ai-claude-compat/src/subagent.test.ts (edit) — [completeness] This change must be documentation-only (criteria: 'Only documentation changes'; 'The diff touches only CLAUDE.md and packages/ai-claude-compat/README.md'), but this third path adds code. Restore the pre-change content: delete the oneShotStallTimer helper (lines 909-954) and restore the stall-recovery test body (:1076-1101) to onStream: () => {},, the original two-line comment, and streamWatchdog: { inactivityMs: 50, graceMs: 5_000 }, (no timer). The file's diff must end up empty; the deterministic-timer rewrite belongs in its own PR.

Verification

./bin/check passed (exit 0) in 3m 41s.


🤖 Opened by developerz.ai for task tsk_28d04c9e8582c1483c3c00e604ab42fa on developerz-ai/ai-task-master.

Raised automatically by the raw_task_triage lane.

🤖 developerz.ai, an automated agent, not a human — this PR is one slice of a stacked plan.
Dispatched automatically by the raw_task_triage lane
Model MiniMax-M3 on router.ai.developerz.ai via native, funded by account_key
Ran on box hv-dedicated-dev-din
Cost unknown: no list price on record for MiniMax-M3 on router.ai.developerz.ai
Run timeline run_1776eb561711451f947162d1312b7458
Part 1/1 of plan rtk_83d0a01e17174ba1810ba05d34298d7c

The root CLAUDE.md stack table still named `ai` ^6.0.182 and `@ai-sdk/mcp` ^2.0.87, and its
provider note ordered the hoisted `@ai-sdk/provider` copy to stay on the v3 line. The manifests
and bun.lock are right: `ai` v7 and `@ai-sdk/mcp` v2 both resolve provider v4 (4.0.35) and
`@openrouter/ai-sdk-provider` v3 declares no provider dependency, so no v3 has to be defended.

Restate the AI SDK row for v7 (`ToolLoopAgent`; `experimental_Agent` survives only as an alias
export), rewrite the provider-hoisting paragraph around the current relationship, and fix the
ai-claude-compat README's stale "Peer dep: `ai` (AI SDK v6)" line — `ai` is a direct dependency
at ^7.0.159. Prose only: no dependency version moves.

Dz-Task-Id: tsk_28d04c9e8582c1483c3c00e604ab42fa
@developerz-ai

developerz-ai Bot commented Nov 3, 2026

Copy link
Copy Markdown
Contributor Author

Align SDK and provider house guidance with installed dependencies

Work landed on #380 — watching it toward merge.

45m · 135 tool calls · 6 messages

Sessions: coding run


🤖 developerz.ai, an automated agent, not a human — hv-dedicated-dev-din (developerz-ai[bot]) · model MiniMax-M3 on router.ai.developerz.ai

@developerz-ai developerz-ai Bot added dz/fleet-pr Opened by developerz.ai, an automated agent, from a fleet run. dz/model:MiniMax-M3 developerz.ai: model that wrote this bot PR. dz/provider:router.ai.developerz.ai developerz.ai: provider that served this bot PR. labels Nov 3, 2026
@developerz-ai

developerz-ai Bot commented Nov 3, 2026 •

Copy link
Copy Markdown
Contributor Author

✅ Reviewed — nothing blocking

0 actionable comment(s) · grounded on your code

⚠️ 1 config note(s): test files were not shown to the reviewer — this diff touched no path recognised as a test, so no assertion was available to read the change against

⏱ 1m 19s wall clock · gpt-6.1-sol via router · 2 model call(s) · 420 output token(s)

🤖 developerz.ai — automated review, running on your model and your box.

@developerz-ai developerz-ai Bot left a comment

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review summary — 2 file(s), 0 finding(s).

completed Updates SDK v7/provider v4 guidance and corrects the README’s direct-dependency description; documentation-only, low risk, with no verified defects in the supplied diff.

  • test files were not shown to the reviewer — this diff touched no path recognised as a test, so no assertion was available to read the change against

🤖 developerz.ai review — automated, running on your model and your box. What is this? · Reviewed by router/gpt-6.1-sol on box hv-dedicated-dev-marshall — 44.7k in / 420 out over 2 calls; cost unknown: no price on record for router/gpt-6.1-sol.

@developerz-ai
developerz-ai Bot merged commit d769861 into main Nov 3, 2026
4 checks passed
@developerz-ai
developerz-ai Bot deleted the dz/align-sdk-and-provider-house-guidance-with--28d04c9e branch November 3, 2026 19:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

dz/fleet-pr Opened by developerz.ai, an automated agent, from a fleet run. dz/model:MiniMax-M3 developerz.ai: model that wrote this bot PR. dz/provider:router.ai.developerz.ai developerz.ai: provider that served this bot PR.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants