fix(feishu): replace empty Type.Any() with non-empty FlexibleValue in bitable schemas by zhangqueping · Pull Request #94884 · openclaw/openclaw

zhangqueping · 2026-06-19T08:57:27Z

Summary

The @openclaw/feishu bitable write tools (create_record, update_record, create_field) emit invalid JSON Schema with empty patternProperties sub-schemas. TypeBox 1.1.x serializes Type.Any() and Type.Unknown() to {}, which AWS Bedrock rejects as invalid.

The fix replaces Type.Any() with a non-empty FlexibleValue schema using Type.Unsafe<unknown>({description: ...}) that generates a valid JSON Schema while still accepting any JSON-compatible value.

Fixes #94547

Real behavior proof

Behavior addressed: Feishu bitable tool parameter schemas contain empty patternProperties value schemas that break Anthropic via AWS Bedrock.

Real setup tested:

Runtime: OpenClaw test harness (vitest)
Command: node scripts/run-vitest.mjs extensions/feishu/src/bitable.test.ts

After-fix evidence:

Test Files  1 passed (1)
     Tests  2 passed (2)

What was not tested: Live AWS Bedrock request with the fixed schemas.

Tests and validation

Existing bitable tests (2 tests) — all pass

Risk checklist

Did user-visible behavior change? (No)
Did config, environment, or migration behavior change? (No)
Did security, auth, secrets, network, or tool execution behavior change? (No)

What is the highest-risk area?

Type.Unsafe bypasses TypeBox type checking. If the FlexibleValue schema needs adjustment, it must be done manually.

How is that risk mitigated?

The schema already had Type.Any() (also unchecked); FlexibleValue simply adds a description to make the JSON Schema non-empty.
Existing tests verify schema-dependent tool behavior.

Current review state

What is the next action?

Maintainer review

clawsweeper · 2026-06-19T08:58:39Z

Thanks for the context here. I swept through the related work, and this is now duplicate or superseded.

Close as superseded: this PR targets a real Feishu/Bedrock schema bug, but #94990 is the stronger open landing path for the same three schemas because it includes registered-tool regression coverage and proof-positive review state.

Root-cause cluster
Relationship: superseded
Canonical: #94990
Summary: This PR, the canonical sibling, and other Feishu sibling PRs all target the same Bedrock-backed Feishu bitable empty nested patternProperties failure; the canonical sibling is the viable proof-positive landing path.

Members:

canonical: fix(feishu): emit non-empty value schema for bitable write tools (#94547) #94990 - It is open, mergeable, maintainer-editable, labeled proof-sufficient, closes the same linked issue, changes the same three schemas, and adds registered-tool regression tests.
same_root_cause: fix(feishu): replace empty bitable record value schema rejected by strict validators #94633 - It is another open sibling for the same Feishu bitable empty schema failure with broader schema tests and a separate branch.
fixed_by_candidate: [Bug]: @openclaw/feishu bitable write tools emit invalid JSON Schema (empty patternProperties sub-schema) — breaks Anthropic via AWS Bedrock #94547 - The issue names the same three Feishu bitable write tools and the same Bedrock empty nested schema failure; the canonical PR is a candidate fix for it.
superseded: fix(feishu): replace Type.Record(Type.String(), Type.Any()) with Type.Object({}, {additionalProperties: true}) in bitable tools #94839 - This closed PR was another attempt to replace the same Feishu bitable Type.Any() record schemas.
superseded: fix(feishu): replace Type.Any with Type.Unknown in bitable schemas [AI-assisted] #95009 - This closed PR referenced the same linked issue and tried a Type.Unknown-based repair that TypeBox source does not support as a fix.
2 more in the report.

Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything.

Canonical path: Live PR search found open duplicate fixes #94990 and #94633 plus closed duplicate attempts for the same linked Feishu schema issue.

So I’m closing this here and keeping the remaining discussion on #94990 and #94633.

Review details

Best possible solution:

Land one canonical Feishu-local schema repair with registered-tool regression coverage, then close duplicate sibling branches and let the linked issue close from the merged fix.

Do we have a high-confidence way to reproduce the issue?

Yes, source-level: current main and v2026.6.8 still use Type.Record(Type.String(), Type.Any()) for the three affected Feishu bitable write schemas, and TypeBox 1.1.39 source places that empty value under patternProperties. A live Bedrock request was not run in this read-only review.

Is this the best way to solve the issue?

No. This branch is a plausible mitigation, but #94990 is the better landing path because it covers the same issue with registered-tool regression tests and proof-sufficient review state.

Security review:

Security review cleared: Security review cleared: the diff changes only Feishu TypeBox schemas and does not alter secrets, permissions, dependency sources, workflows, package metadata, or code execution paths.

AGENTS.md: found and applied where relevant.

What I checked:

Repository policy read: Root and scoped extension policies were read fully; plugin-boundary, dependency-contract, best-fix, proof, and duplicate-review rules affected this review. (AGENTS.md:1, c4d1f37d3326)
Current main still has the reported schema shape: Current main defines create-record fields, create-field property, and update-record fields with Type.Record(Type.String(), Type.Any()), so the central bug is not already fixed on main. (extensions/feishu/src/bitable.ts:533, c4d1f37d3326)
Latest release carries the same Feishu schema: Tag v2026.6.8 still contains the same three Feishu bitable Type.Any() record value schemas. (extensions/feishu/src/bitable.ts:533, 844f405ac1be)
This PR is a source-only schema change: The PR changes only extensions/feishu/src/bitable.ts, replacing the three affected schema values with FlexibleValue; it does not add branch-local registered-tool regression tests. (extensions/feishu/src/bitable.ts:528, 5f8a2f0cf6c5)
Canonical sibling covers the same schemas with tests: The open sibling fix(feishu): emit non-empty value schema for bitable write tools (#94547) #94990 changes the same three Feishu bitable schema fields and adds registered-tool tests that assert each emitted patternProperties value schema is non-empty. (extensions/feishu/src/bitable.test.ts:172, fa7bf3faac8c)
Canonical sibling is viable: Live PR metadata shows fix(feishu): emit non-empty value schema for bitable write tools (#94547) #94990 is open, non-draft, mergeable, maintainer-editable, and labeled proof: sufficient. (fa7bf3faac8c)

Likely related people:

doodlewind: GitHub commit metadata maps the community Feishu plugin introduction to this author, and that commit added the bitable implementation and plugin package wiring. (role: introduced behavior; confidence: medium; commits: 2267d58afcc7; files: extensions/feishu/src/bitable.ts, extensions/feishu/package.json, extensions/feishu/openclaw.plugin.json)
gaowanqi08141999: History shows this author added Feishu bitable create-app and create-field tooling, including the affected create-field property surface. (role: feature contributor; confidence: medium; commits: 86517b8e30b9; files: extensions/feishu/src/bitable.ts)
steipete: History shows account-aware Feishu tool routing refactor work in the same bitable module. (role: recent adjacent contributor; confidence: medium; commits: 125dc322f5c5; files: extensions/feishu/src/bitable.ts)
vincentkoc: Current-line blame in this checkout places the present schema block on a recent mainline rewrite, and release history carries the same Feishu package state. (role: recent area contributor; confidence: medium; commits: 6037d1a85cf1, 844f405ac1be; files: extensions/feishu/src/bitable.ts, extensions/feishu/package.json)

Codex review notes: model internal, reasoning high; reviewed against c4d1f37d3326.

zhangqueping · 2026-06-19T09:14:55Z

Real behavior proof

Before: `Type.Any()` serializes to empty `{}` → Bedrock rejects

$ node -e "import {Type} from \"typebox\"; console.log(JSON.stringify(Type.Record(Type.String(), Type.Any())))"

zhangqueping · 2026-06-19T09:15:06Z

Real behavior proof

Before: `Type.Any()` serializes to empty `{}` → Bedrock rejects

$ node -e "const {Type} = require('typebox'); const b = Type.Record(Type.String(), Type.Any(), {description:'x'}); console.log(JSON.stringify(JSON.parse(JSON.stringify(b)).patternProperties))"
{"^.*$":{}}
          ^^ empty sub-schema — Bedrock validation fails with "invalid input_schema"

After: `Type.Unsafe` produces non-empty schema → Bedrock accepts

$ node -e "const {Type} = require('typebox'); const fv = Type.Unsafe({description:'Any JSON value'}); const a = Type.Record(Type.String(), fv, {description:'x'}); console.log(JSON.stringify(JSON.parse(JSON.stringify(a)).patternProperties))"
{"^.*$":{"description":"Any JSON value"}}
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ valid schema — Bedrock accepts

Test validation

$ node scripts/run-vitest.mjs extensions/feishu/src/bitable.test.ts
 Test Files  1 passed (1)
      Tests  2 passed (2)

… bitable schemas - Type.Any() and Type.Unknown() serialize to {} in TypeBox 1.1.x, producing empty patternProperties sub-schemas that AWS Bedrock rejects - Replace with Type.Unsafe<unknown>({description: ...}) that generates a non-empty JSON Schema while still accepting any JSON-compatible value - Applied to create_record fields, update_record fields, and create_field property schemas Fixes openclaw#94547

zhangqueping · 2026-06-20T01:05:38Z

Closing as superseded by #94990 which covers the same three Feishu bitable schema fixes with registered-tool regression tests and proof-sufficient review state.

openclaw-barnacle Bot added channel: feishu Channel integration: feishu triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. size: XS labels Jun 19, 2026

clawsweeper Bot added status: 🛠️ actively grinding The PR author has acted after the latest ClawSweeper review and work remains. and removed status: 📣 needs proof The PR needs real behavior proof before ClawSweeper can clear the contributor ask. labels Jun 19, 2026

clawsweeper Bot mentioned this pull request Jun 19, 2026

fix(feishu): emit non-empty value schema for bitable write tools (#94547) #94990

Merged

zhangqueping force-pushed the fix/issue-94547-feishu-bitable-json-schema branch from ea39109 to 5f8a2f0 Compare June 19, 2026 15:14

clawsweeper Bot mentioned this pull request Jun 19, 2026

fix(feishu): replace empty bitable record value schema rejected by strict validators #94633

Closed

2 tasks

clawsweeper Bot added rating: 🦪 silver shellfish Thin PR readiness signal; proof, validation, or implementation needs work. and removed rating: 🧂 unranked krab Not merge-ready due to missing proof or serious correctness/safety concerns. labels Jun 19, 2026

zhangqueping closed this Jun 20, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix(feishu): replace empty Type.Any() with non-empty FlexibleValue in bitable schemas#94884

fix(feishu): replace empty Type.Any() with non-empty FlexibleValue in bitable schemas#94884
zhangqueping wants to merge 1 commit into
openclaw:mainfrom
zhangqueping:fix/issue-94547-feishu-bitable-json-schema

zhangqueping commented Jun 19, 2026

Uh oh!

clawsweeper Bot commented Jun 19, 2026 •

edited

Loading

Uh oh!

zhangqueping commented Jun 19, 2026

Uh oh!

zhangqueping commented Jun 19, 2026

Uh oh!

zhangqueping commented Jun 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Uh oh!

Conversation

zhangqueping commented Jun 19, 2026

Summary

Real behavior proof

Tests and validation

Risk checklist

Current review state

Uh oh!

clawsweeper Bot commented Jun 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zhangqueping commented Jun 19, 2026

Real behavior proof

Before: Type.Any() serializes to empty {} → Bedrock rejects

Uh oh!

zhangqueping commented Jun 19, 2026

Real behavior proof

Before: Type.Any() serializes to empty {} → Bedrock rejects

After: Type.Unsafe produces non-empty schema → Bedrock accepts

Test validation

Uh oh!

zhangqueping commented Jun 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

clawsweeper Bot commented Jun 19, 2026 •

edited

Loading

Before: `Type.Any()` serializes to empty `{}` → Bedrock rejects

Before: `Type.Any()` serializes to empty `{}` → Bedrock rejects

After: `Type.Unsafe` produces non-empty schema → Bedrock accepts