@tangle-network/agent-knowledge 13.0.1 → 14.0.1

package/docs/architecture.md

@@ -58,6 +58,10 @@ After the exact submitted bytes are durable, `runVerifiedResearchLoop` passes th
 The ledger materializes only observations whose complete source identity matches a confirmed record, so reusing one URI for different bytes cannot activate the wrong claims and a crash on either side resumes safely.
 Unversioned URI-only ledgers cannot prove which bytes produced their observations; reads and writes fail with `ClaimLedgerMigrationRequiredError` and preserve the original file for an explicit archive-and-reverify migration.
 Before synchronous question generation, the persistent driver records `preparedRounds`; a resume reconstructs and checkpoints any prepared round whose questions were interrupted, and the loop publishes its `research.iteration` event only after that checkpoint succeeds.
+The research loop requires storage readiness and the driver's optional `isComplete()` result before it reports completion.
+An unfinished driver can generate steering with no remaining storage gaps, so passing source requirements does not stop research prematurely.
+Drivers without `isComplete()` use storage readiness alone.
+Without readiness specifications, the loop runs to its round limit and never reports ready.
 
 Every write in this layer goes through `durable-fs` (`writeFileDurable`, `writeJsonDurableWithinRoot`): temp file, fsync, atomic rename, and parent fsync.
 `O_NOFOLLOW` descriptors anchored through `/proc/self/fd` prevent a directory swapped for a symlink during a write from redirecting it outside the root.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tangle-network/agent-knowledge",
-  "version": "13.0.1",
+  "version": "14.0.1",
   "description": "Build, search, evaluate, and improve source-backed knowledge bases.",
   "homepage": "https://github.com/tangle-network/agent-knowledge#readme",
   "repository": {
@@ -84,14 +84,14 @@
     "zod": "4.5.4"
   },
   "peerDependencies": {
-    "@tangle-network/agent-eval": ">=0.173.0 <0.174.0",
+    "@tangle-network/agent-eval": ">=0.174.0 <0.175.0",
     "@tangle-network/agent-interface": "^2.0.0"
   },
   "devDependencies": {
     "@arethetypeswrong/cli": "^0.18.5",
     "@biomejs/biome": "^2.5.11",
     "@neo4j-labs/agent-memory": "0.4.1",
-    "@tangle-network/agent-eval": "0.173.0",
+    "@tangle-network/agent-eval": "0.174.0",
     "@tangle-network/agent-interface": "2.0.0",
     "@types/node": "^26.4.0",
     "mem0ai": "3.1.7",

package/skills/build-with-agent-knowledge/SKILL.md

@@ -56,6 +56,10 @@ Use existing vector, graph, search, and memory systems through adapters instead
 3. Run one representative query or memory sequence through the production path.
 4. Capture retrieved items, final answer or action, citations, errors, tokens, cost, and latency.
 5. Prove a known good case passes and a realistic unsupported or missed case fails.
+
+For ingestion or retrieval work, stop when the requested path passes these checks.
+When the task requires improving an existing result:
+
 6. Add only the missing improvement step: retrieval search, source acquisition, page update, answer repair, or memory policy.
 7. Write changes to an isolated candidate with a stable run identity.
 8. Compare baseline and candidate on the same development cases, then on unseen cases.
@@ -79,19 +83,14 @@ Bundled benchmark samples prove adapter wiring only; use complete external datas
 
 ## Completion
 
-One customer-like path must prove:
-
-```text
-source or memory event -> production retrieval -> observable answer or action
--> isolated candidate -> baseline comparison -> unseen comparison
--> approved promotion or correctly rejected change -> reproducible rerun
-```
-
-Report installed versions, exact imports, provider adapters, scope policy, case counts, baseline and candidate results, cost, latency, candidate identity, promotion result, and artifact paths.
+Prove the requested source or memory event reaches production retrieval and an observable answer or action.
+For improvement work, also prove isolation, baseline and unseen comparisons, and approved promotion or correct rejection.
+Report the checked path, scope policy, results, material limits, and reproducible evidence.
+Include dependency identity only when it explains compatibility or the result.
 
 ## Then consider
 
 - `eval-engineering` when new production-derived cases are needed.
 - `build-with-agent-runtime` when agents should research, edit, or compare candidates.
-- `agent-eval-adoption` when the product needs shared comparison and release records.
+- `agent-eval` when the product needs shared comparison and release records.
 - `harden` when changing tenant isolation, source trust, deletion, or promotion authority.