Hotfix: separate workspace observation from lock claim — restore artifact projector for cognitive runs

Production regression of T1 (7829acc..15e3863). User reports "明明生成
了很多文件,但只有一个" — the filesystem shows 19 markdown deliverables
in ~/Documents/Github/test/{business-plan,operations,ceo-coordination}/
but the DocumentCenter surfaces only one row.

Root cause is structural, predicted in spirit by Karpathy's Round 1
take ("把'认知'和'执行'两件事一起锁是数据结构错"). The T1 rewrite
collapsed two concerns into one nullable field on the run context:

  workspacePath = workspace?.localPath ?? null
                  ^^^^^^^^^                ^^^^
                  lock target              observation root

For lockMode='none' (every cognitive C-suite agent post-T1) the
resolver returns workspace=null, so workspacePath was null, so
artifact-projector.ts:242 short-circuited every projection with
`{ kind: "skipped", reason: "no_workspace" }`. The agents kept
writing files via bash (cwd falls back to the company default
workspace at worker/runtime/index.ts:142), but nothing observed
them — the DocumentCenter only saw the one file that was projected
back when the agent still held an exclusive workspace.

Fix: split the two concerns. WorkspaceResolution gains a non-lock
`observationPath` field; the resolver populates it from the company
primary workspace (new `getCompanyObservationWorkspace` query —
primary first, oldest second, null only when company has zero
workspace rows) for lockMode='none' runs, and from `workspace.localPath`
itself for lockMode='exclusive' runs. RunDispatchParams threads the
field through; turn-context's `workspacePath` reads it first.
artifact-projector.ts is unchanged — it just sees a non-null path
again and resumes baselining, diffing, and upserting work_products
rows the way it did pre-T1.

Three new unit tests pin the regression: cognitive role with
workspace ⇒ observationPath set; cognitive role with no workspace ⇒
observationPath null; exclusive ⇒ observationPath equals
workspace.localPath. typecheck green, 543/543 unit tests, 10/10
eval. No fallback, no mock, no silent degradation.
This commit is contained in:
ZacharyZhang-NY committed 2026-04-20 16:22:40 -04:00
1 parent e15bf67c9f
commit 1f2f9770a6
6 files changed
+132 -4

No files matched your search

+22
View File
@@ -61,6 +61,28 @@ export function buildWorkspaceMethods(
return row ? { ...row, isPrimary: toBoolean(row.isPrimary) } : null;
},
/**
* Pick a single workspace per company for *observation only* (no
* lock claim). Used by the artifact projector to know where on
* disk a `lockMode='none'` cognitive run might write files. The
* tie-break order is `is_primary` first, then earliest `created_at`
* — i.e. the workspace the operator marked as canonical wins, and
* if no one is marked, the oldest row wins (which historically is
* the company's default scratch directory).
*
* Returns `null` when the company has no workspace at all; the
* caller treats that as "agent has no observation surface", which
* is correct — there is genuinely nowhere to look for files.
*/
getCompanyObservationWorkspace(companyId: string): WorkspaceRecord | null {
const row = getDb()
.prepare(
"select id, company_id as companyId, name, local_path as localPath, repo_url as repoUrl, repo_ref as repoRef, project_id as projectId, state, is_primary as isPrimary, created_at as createdAt, updated_at as updatedAt from workspaces where company_id = ? order by is_primary desc, created_at asc limit 1",
)
.get(companyId) as WorkspaceRecord | undefined;
return row ? { ...row, isPrimary: toBoolean(row.isPrimary) } : null;
},
listWorkspaceLocalPaths(companyId: string): string[] {
const rows = getDb()
.prepare("select local_path as localPath from workspaces where company_id = ? and local_path is not null")
+1
View File
@@ -255,6 +255,7 @@ export function createHeartbeat(
connector,
workspace,
workspaceMode: workspaceResolution.lockMode,
observationPath: workspaceResolution.observationPath,
companyName: companyBasic?.name ?? "Unknown",
companyDescription: companyBasic?.description ?? "",
snapshot: db.snap.listSnapshot(),
+44 -2
View File
@@ -49,9 +49,33 @@ export type WorkspaceResolutionSource =
| "none";
export interface WorkspaceResolution {
/**
* The workspace whose lock the dispatcher will claim. Non-null iff
* `lockMode === 'exclusive'` — the type system enforces the pair.
*/
workspace: WorkspaceRecord | null;
lockMode: WorkspaceLockMode;
source: WorkspaceResolutionSource;
/**
* Where the artifact projector should *observe* for files this run
* may produce, independent of the lock decision. Set whenever the
* company has at least one workspace row, even when `lockMode='none'`
* — cognitive agents still write deliverables to disk via `bash`,
* and the projector needs an observation root to scan against the
* mtime baseline. Null only when the company has no workspace at
* all (genuinely nowhere to look).
*
* For `lockMode='exclusive'` this is always the same path as
* `workspace.localPath`. For `lockMode='none'` this is the
* company's primary/oldest workspace localPath, picked by
* `getCompanyObservationWorkspace` — *no lock claim*.
*
* Decoupling observation from locking is what fixes the regression
* where cognitive agents wrote 19 files but the DocumentCenter
* surfaced only 1: previously `workspacePath` shared the lock
* decision, so any non-exclusive run lost its projector.
*/
observationPath: string | null;
}
export class WorkspaceConfigurationError extends Error {
@@ -177,7 +201,19 @@ export function resolveWorkspaceWithLockMode(
const decision = inferLockMode(db, task, agent, companyId);
if (decision.lockMode === "none") {
return { workspace: null, lockMode: "none", source: "none" };
// No lock claim, but the cognitive run can still touch the
// filesystem via `bash`. Pick a stable observation root so the
// artifact projector can baseline + diff the run's footprint.
// `getCompanyObservationWorkspace` returns the primary or oldest
// workspace; null only when the company has no workspace row at
// all, in which case there is genuinely nothing to observe.
const observation = db.workspacesMod.getCompanyObservationWorkspace(companyId);
return {
workspace: null,
lockMode: "none",
source: "none",
observationPath: observation?.localPath ?? null,
};
}
if (!decision.workspaceId) {
@@ -198,5 +234,11 @@ export function resolveWorkspaceWithLockMode(
);
}
return { workspace, lockMode: "exclusive", source: decision.source };
return {
workspace,
lockMode: "exclusive",
source: decision.source,
// Exclusive runs observe the same path they lock — single source.
observationPath: workspace.localPath,
};
}
+2
View File
@@ -130,6 +130,7 @@ export function createTaskRunner(
connector,
workspace,
workspaceMode: workspaceResolution.lockMode,
observationPath: workspaceResolution.observationPath,
companyName: companyBasic?.name ?? "Unknown",
companyDescription: companyBasic?.description ?? "",
snapshot: db.snap.listSnapshot(),
@@ -230,6 +231,7 @@ export function createTaskRunner(
connector,
workspace: workspaceResolution.workspace,
workspaceMode: workspaceResolution.lockMode,
observationPath: workspaceResolution.observationPath,
companyName: company.name,
companyDescription: company.description,
snapshot,
+18 -2
View File
@@ -28,6 +28,16 @@ export interface RunDispatchParams {
connector: ReturnType<AppDatabase["connectorsCrud"]["getConnector"]>;
workspace: ReturnType<AppDatabase["workspacesMod"]["getWorkspace"]>;
workspaceMode: WorkspaceLockMode;
/**
* Filesystem root the artifact projector observes for files this run
* may produce, independent of `workspaceMode`. Set whenever the
* resolver found a workspace row in the company (including for
* `lockMode='none'` cognitive runs that do not claim the lock but
* still write deliverables via `bash`). Optional so legacy callers
* that bypass the resolver keep typechecking; the new resolver
* always sets it.
*/
observationPath?: string | null;
companyName: string;
companyDescription: string;
snapshot: ReturnType<AppDatabase["snap"]["listSnapshot"]>;
@@ -286,7 +296,7 @@ function buildConnectorCatalog(
* This is the core composition step that was previously inlined in dispatchRunToWorker.
*/
export function buildRunPayload(params: RunDispatchParams, deps: TurnContextDeps): EnqueueRunPayload {
const { runId, companyId, agent, task, connector, workspace, workspaceMode, companyName, companyDescription, snapshot, trigger } = params;
const { runId, companyId, agent, task, connector, workspace, workspaceMode, observationPath, companyName, companyDescription, snapshot, trigger } = params;
const { db } = deps;
const { env, apiUrl, apiKey } = assembleApiEnv(db, connector, companyId, agent.id, runId, trigger, deps);
@@ -375,7 +385,13 @@ export function buildRunPayload(params: RunDispatchParams, deps: TurnContextDeps
agentName: agent.name,
connectorId: connector.id,
workspaceId: workspaceMode === "exclusive" ? (workspace?.id ?? null) : null,
workspacePath: workspace?.localPath ?? null,
// observationPath, not lock target. Cognitive runs (lockMode='none')
// still write deliverables via bash; the artifact projector needs an
// observation root regardless of whether we hold the lock. Falls back
// to workspace?.localPath for old call sites that did not yet thread
// the observationPath through (the resolver always sets it now;
// callers that bypass the resolver remain on the legacy field).
workspacePath: observationPath ?? workspace?.localPath ?? null,
workspaceMode,
model: modelSelection.effectiveModel,
thinkingEffort,
+45
View File
@@ -217,4 +217,49 @@ describe("resolveWorkspaceWithLockMode", () => {
expect(m.workspace).toBeNull();
}
});
it("lockMode=none still returns observationPath when company has a workspace (artifact projector regression)", () => {
const { companyId } = seedCompany(db);
const result = resolveWorkspaceWithLockMode(
db,
null,
{ workspaceId: null, role: "CMO", department: "marketing" },
companyId,
);
expect(result.lockMode).toBe("none");
expect(result.workspace).toBeNull();
// Observation path is the company's primary workspace localPath
// even though no lock is claimed — the artifact projector needs
// somewhere to baseline + diff so files written by cognitive
// runs surface in the DocumentCenter.
expect(result.observationPath).toBe("/tmp/fixture");
});
it("lockMode=none observationPath is null when company has no workspace at all", () => {
const companyId = db.company.saveCompany({
name: "No-workspace Co",
description: "",
status: "active",
});
const result = resolveWorkspaceWithLockMode(
db,
null,
{ workspaceId: null, role: "CFO", department: "finance" },
companyId,
);
expect(result.lockMode).toBe("none");
expect(result.observationPath).toBeNull();
});
it("lockMode=exclusive observationPath equals workspace.localPath (single source)", () => {
const { companyId, workspaceId } = seedCompany(db);
const result = resolveWorkspaceWithLockMode(
db,
null,
{ workspaceId, role: "CTO", department: "engineering" },
companyId,
);
expect(result.lockMode).toBe("exclusive");
expect(result.observationPath).toBe(result.workspace?.localPath);
});
});