@cairnkeep/cli 2.21.0 → 2.21.1

package/CHANGELOG.md

@@ -5,6 +5,22 @@ All notable user-facing changes are documented here. This project follows
 
 ## [Unreleased]
 
+## [2.21.1] - 2026-10-07
+
+### Fixed
+
+- Classify embedding health failures by sanitized HTTP status so operators can
+  distinguish authorization, missing endpoint/model, throttling, transport,
+  and upstream failures without exposing provider response bodies, request
+  identifiers, or credentials.
+
+### CI
+
+- Split all 48 memory-server smoke contracts exactly once across three
+  deterministically balanced native-Windows jobs while retaining Windows CLI,
+  package, global-install, Node 22/24/26, and full Linux matrix coverage. This
+  removes the serial 15-minute suite duplication from every Windows lane.
+
 ## [2.21.0] - 2026-10-07
 
 ### Added

package/docs/learning/CURRICULUM-MAP.md

@@ -1,6 +1,6 @@
 # Curriculum Coverage Map
 
-**Baseline:** Cairnkeep 2.21.0
+**Baseline:** Cairnkeep 2.21.1
 **Last reviewed:** 2026-10-06
 
 This map prevents the course from drifting behind the product. It assigns each

package/docs/learning/FEATURE-GUIDE.md

@@ -1,6 +1,6 @@
 # Cairnkeep feature guide
 
-**Baseline:** Cairnkeep 2.21.0
+**Baseline:** Cairnkeep 2.21.1
 
 This is the step-back map for choosing and operating Cairnkeep features. The
 [course labs](https://github.com/cairnkeep/cairnkeep-course-labs) provide one

package/docs/learning/lessons/L00-why-cairnkeep.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Quickstart
 **Time:** 10 minutes
-**Tested with:** Cairnkeep 2.21.0
+**Tested with:** Cairnkeep 2.21.1
 
 ## Outcome
 

package/docs/learning/lessons/L01-safe-trial.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Quickstart
 **Time:** 15 minutes
-**Tested with:** Cairnkeep 2.21.0
+**Tested with:** Cairnkeep 2.21.1
 
 ## Outcome
 

package/docs/learning/lessons/L02-installation.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Quickstart
 **Time:** 25 minutes
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 
 ## Outcome
 

package/docs/learning/lessons/L03-first-project.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Quickstart
 **Time:** 25 minutes
-**Tested with:** Cairnkeep 2.21.0
+**Tested with:** Cairnkeep 2.21.1
 
 ## Outcome
 

package/docs/learning/lessons/L19-validated-skill-improvement.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Evidence and Evaluation
 **Time:** 50 minutes
-**Tested with:** Cairnkeep 2.21.0
+**Tested with:** Cairnkeep 2.21.1
 
 ## Outcome
 

package/docs/learning/lessons/L20-mcp-tool-profiles.md

@@ -1,7 +1,7 @@
 # L20 - Least-authority MCP tool profiles
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 **Time:** 25 minutes
 
 ## Outcome

package/docs/learning/lessons/L21-context-packs.md

@@ -1,7 +1,7 @@
 # L21 - Immutable context packs and skill approval
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 **Time:** 40 minutes
 
 ## Outcome

package/docs/learning/lessons/L22-native-windows.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Operator
 **Time:** 30 minutes
-**Tested with:** Cairnkeep 2.21.0, Windows x64, and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1, Windows x64, and Node.js 22 or newer
 
 ## Outcome
 

package/docs/learning/lessons/L23-guided-setup.md

@@ -3,7 +3,7 @@
 **Status:** Ready
 **Track:** Quickstart and Operator
 **Time:** 35 minutes
-**Tested with:** Cairnkeep 2.21.0; Pi 0.84.1; Node.js 22, 24, and 26
+**Tested with:** Cairnkeep 2.21.1; Pi 0.84.1; Node.js 22, 24, and 26
 
 ## Outcome
 

package/docs/learning/lessons/L24-okf-exchange.md

@@ -1,7 +1,7 @@
 # L24 - Reviewed knowledge exchange
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 **Time:** 35 minutes
 
 ## Outcome

package/docs/learning/lessons/L25-playbooks.md

@@ -1,7 +1,7 @@
 # L25 - Bounded workflow playbooks
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 **Time:** 35 minutes
 
 ## Outcome

package/docs/learning/lessons/L26-context-intelligence.md

@@ -1,7 +1,7 @@
 # L26 - Context intelligence with explicit authority
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 
 ## Outcome
 

package/docs/learning/lessons/L27-ecosystem-evidence.md

@@ -1,6 +1,6 @@
 # L27 - Evaluate context companions without moving authority
 
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 
 ## Outcome
 

package/docs/learning/lessons/L28-security-assurance.md

@@ -1,7 +1,7 @@
 # L28 - Verify Cairnkeep's security posture
 
 **Status:** Ready
-**Tested with:** Cairnkeep 2.21.0 and Node.js 22 or newer
+**Tested with:** Cairnkeep 2.21.1 and Node.js 22 or newer
 
 ## Outcome
 
@@ -11,7 +11,7 @@ and choose least authority for a network-facing MCP client.
 
 ## Prerequisites
 
-- Install Cairnkeep 2.21.0 or use a disposable source checkout.
+- Install Cairnkeep 2.21.1 or use a disposable source checkout.
 - Complete [L20 - Least-authority MCP tool profiles](L20-mcp-tool-profiles.md).
 - Use synthetic values only; never paste a real bearer token into a recording,
   terminal transcript, or issue.

package/docs/operating.md

@@ -1049,6 +1049,10 @@ then checks `./.ai/.env` (or the current environment). It does not inspect a
 harness's remote HTTP registration. Unconfigured optional dependencies are
 skipped; it exits non-zero when the local server probe fails or a configured
 dependency (LLM/embedding endpoint, writable store) is unreachable.
+Embedding probes report only a bounded classification and HTTP status, such as
+`authorization rejected (HTTP 403)` or `endpoint or model not found (HTTP
+404)`. Provider response bodies and request identifiers are discarded so a
+health check cannot echo sensitive upstream diagnostics.
 
 It also checks an existing project-local trajectory database for SQLite
 integrity, schema compatibility and index consistency. Because capture is

package/docs/releasing.md

@@ -29,6 +29,13 @@ from a workstation.
 9. Verify both image digests, SBOM attestations, provenance attestations, and an
    anonymous pull before announcing the release.
 
+The native-Windows lanes retain stable check names for branch protection. Each
+lane exercises the Windows CLI and packed global install on its Node version,
+while the complete memory-server smoke catalog is deterministically balanced
+across the three lanes. The Linux Node 22/24/26 matrix independently runs the
+complete catalog on every supported runtime. This preserves coverage while
+removing three serial copies of the slow Windows suite from the critical path.
+
 Stable versions must use a stable GitHub Release and publish to npm's `latest`
 tag. SemVer prerelease versions must use a GitHub prerelease and publish to
 `next`. A mismatch fails before publication.

package/docs/security-assurance.md

@@ -86,6 +86,10 @@ punctuation.
 `cairn doctor` remains the operational dependency/store check. Run both before
 exposing HTTP, after changing tool authority, and before a security-sensitive
 release.
+Its embedding probe retains only a sanitized failure class and HTTP status;
+provider response bodies and request identifiers are never emitted. This makes
+expired credentials distinguishable from a missing model without turning the
+doctor output into a secondary secret or telemetry channel.
 
 ## Repository and release controls
 
@@ -120,3 +124,7 @@ The maintained gates are `npm run security:baseline`, both production
 security-assurance and HTTP-guard smoke tests, native-Windows tests, and the
 complete public test matrix. The baseline gate also rejects
 `pull_request_target` and every unpinned non-local GitHub Action.
+The native-Windows matrix runs the CLI, package, and global-install contracts on
+Node 22, 24, and 26. It deterministically partitions the complete memory-server
+smoke catalog across those lanes exactly once; the Linux matrix still runs the
+complete catalog independently on every supported Node version.

package/examples/eval/bundled-fake.json

@@ -1 +1 @@
-{"identifier":"cairn-offline-fake-v1","package_version":"2.21.0","schema_version":1,"task_set_digest":"13e95379f35839593047ec34521a65a8fa3ab92bd3993bca9dc8bf5c4a9ff36f"}
+{"identifier":"cairn-offline-fake-v1","package_version":"2.21.1","schema_version":1,"task_set_digest":"13e95379f35839593047ec34521a65a8fa3ab92bd3993bca9dc8bf5c4a9ff36f"}

package/mcp-memory-server/package.json

@@ -62,7 +62,7 @@
         "check:security-assurance": "node scripts/smoke-security-assurance.mjs",
         "check:notes": "npm run build && npm run check:note-signatures && npm run check:note-distillation && npm run check:note-enrichment",
         "test": "npm run test:smoke",
-        "test:smoke": "npm run build && npm run check:embeddings && npm run check:extract && npm run check:scope-guard && npm run check:http-guard && npm run check:remote-context && npm run check:explore-guard && npm run check:explore-cache && npm run check:explore-crossref && npm run check:route-guard && npm run check:reviewed-memory && npm run check:memory-proposals && npm run check:trajectory-roundtrip && npm run check:trajectory-redaction && npm run check:trajectory-retention && npm run check:note-signatures && npm run check:note-distillation && npm run check:note-enrichment && npm run check:node-compat && npm run check:typed-nodes && npm run check:memory-import && npm run check:note-mcp && npm run check:node-doctor && npm run check:artifact-store && npm run check:artifact-mcp && npm run check:compaction-capture && npm run check:capability-contract && npm run check:capability-logging && npm run check:capability-harness && npm run check:capability-mcp && npm run check:eval-schema && npm run check:eval-process && npm run check:eval-statistics && npm run check:skill-lifecycle && npm run check:graph-cli && npm run check:mcp-trust && npm run check:pi-mcp-bridge && npm run check:pi-bridge-env && npm run check:pi-bridge-child && npm run check:context-pack && npm run check:context-pack-retrieval && npm run check:context-pack-v216-compat && npm run check:context-usage && npm run check:domain-retrieval && npm run check:retrieval-benchmark && npm run check:work-evidence && npm run check:okf && npm run check:playbook && npm run check:security-assurance",
+        "test:smoke": "npm run build && node scripts/run-smoke-suite.mjs",
         "test:search": "npm run build && npm run check:search",
         "test:memory": "npm run build && npm run check:embeddings && npm run check:extract && npm run check:search",
         "test:all": "npm run test:memory",

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@cairnkeep/cli",
-    "version": "2.21.0",
+    "version": "2.21.1",
     "description": "A durable, harness-agnostic memory + context layer for coding agents and MCP clients.",
     "license": "Apache-2.0",
     "type": "module",

package/scripts/doctor.sh

@@ -36,15 +36,31 @@ fail() { printf '  [FAIL] %s\n' "$1"; fails=$((fails + 1)); }
 # reachable); a connect/timeout failure returns non-zero.
 reachable() { curl -sS -m 5 -o /dev/null "$1" >/dev/null 2>&1; }
 
+embedding_failure="request failed"
 embedding_works() {
-  local payload
+  local payload http_status curl_status
   payload=$(node -e 'process.stdout.write(JSON.stringify({ model: process.argv[1], input: ["cairnkeep health check"] }))' \
     "$CAIRN_MEMORY_EMBEDDING_MODEL") || return 1
-  curl -fsS -m 10 -o /dev/null \
+  http_status=$(curl -sS -m 10 -o /dev/null -w '%{http_code}' \
     -H "Authorization: Bearer $CAIRN_LLM_API_KEY" \
     -H "Content-Type: application/json" \
     --data "$payload" \
-    "${CAIRN_MEMORY_EMBEDDING_URL%/}/embeddings" >/dev/null 2>&1
+    "${CAIRN_MEMORY_EMBEDDING_URL%/}/embeddings" 2>/dev/null)
+  curl_status=$?
+  if [[ $curl_status -ne 0 ]]; then
+    embedding_failure="transport failed"
+    return 1
+  fi
+  case "$http_status" in
+    2??) return 0 ;;
+    401|403) embedding_failure="authorization rejected (HTTP $http_status)" ;;
+    404) embedding_failure="endpoint or model not found (HTTP 404)" ;;
+    408|429) embedding_failure="temporarily unavailable (HTTP $http_status)" ;;
+    5??) embedding_failure="upstream service failed (HTTP $http_status)" ;;
+    [0-9][0-9][0-9]) embedding_failure="request failed (HTTP $http_status)" ;;
+    *) embedding_failure="request failed" ;;
+  esac
+  return 1
 }
 
 echo "cairn doctor"
@@ -90,7 +106,7 @@ if [[ -n "${CAIRN_MEMORY_EMBEDDING_URL:-}" ]]; then
   elif embedding_works; then
     pass "embedding endpoint accepted model ${CAIRN_MEMORY_EMBEDDING_MODEL} (${CAIRN_MEMORY_EMBEDDING_URL})"
   else
-    fail "embedding request failed for model ${CAIRN_MEMORY_EMBEDDING_MODEL} (${CAIRN_MEMORY_EMBEDDING_URL})"
+    fail "embedding ${embedding_failure} for model ${CAIRN_MEMORY_EMBEDDING_MODEL} (${CAIRN_MEMORY_EMBEDDING_URL})"
   fi
 else
   skip "embedding endpoint (CAIRN_MEMORY_EMBEDDING_URL unset)"

package/scripts/test-ci-sharding.sh

@@ -0,0 +1,33 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT=$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)
+node --input-type=module - "$ROOT" <<'NODE'
+import assert from "node:assert/strict";
+import { readFileSync } from "node:fs";
+import { join } from "node:path";
+import { pathToFileURL } from "node:url";
+
+const root = process.argv[2];
+const { buildShardPlan, smokeChecks } = await import(
+  pathToFileURL(join(root, "mcp-memory-server", "scripts", "run-smoke-suite.mjs")));
+const plan = buildShardPlan(3);
+const flattened = plan.flatMap(({ checks }) => checks.map(({ name }) => name));
+assert.equal(flattened.length, smokeChecks.length);
+assert.equal(new Set(flattened).size, smokeChecks.length);
+assert.deepEqual([...flattened].sort(), smokeChecks.map(({ name }) => name).sort());
+assert.ok(plan.every(({ checks }) => checks.length > 0));
+const weights = plan.map(({ weight }) => weight);
+assert.ok(Math.max(...weights) - Math.min(...weights) <= 5,
+  `native Windows shards are imbalanced: ${weights.join(", ")}`);
+
+const pkg = JSON.parse(readFileSync(join(root, "mcp-memory-server", "package.json"), "utf8"));
+assert.match(pkg.scripts["test:smoke"], /run-smoke-suite\.mjs/);
+const workflow = readFileSync(join(root, ".github", "workflows", "ci.yml"), "utf8");
+assert.match(workflow, /name: native-windows \(\$\{\{ matrix\.node \}\}\)/);
+for (const shard of ["1/3", "2/3", "3/3"]) assert.ok(workflow.includes(`shard: "${shard}"`));
+assert.match(workflow, /run-smoke-suite\.mjs --shard \$\{\{ matrix\.shard \}\}/);
+assert.doesNotMatch(workflow.match(/  native-windows:[\s\S]*?\n  repository:/)?.[0] ?? "",
+  /npm --prefix mcp-memory-server test/);
+console.log(`PASS: ${smokeChecks.length} memory-server contracts split once across balanced Windows shards (${weights.join(", ")})`);
+NODE

package/scripts/test-doctor.sh

@@ -92,13 +92,15 @@ const server = createServer((request, response) => {
   request.on("end", () => {
     let body = {};
     try { body = JSON.parse(Buffer.concat(chunks).toString("utf8")); } catch {}
-    const valid = request.url === "/v1/embeddings"
-      && request.headers.authorization === "Bearer test-key"
-      && body.model === "test-embedding";
-    response.writeHead(valid ? 200 : 404, { "Content-Type": "application/json" });
-    response.end(JSON.stringify(valid
+    const routeValid = request.url === "/v1/embeddings";
+    const keyValid = request.headers.authorization === "Bearer test-key";
+    const modelValid = body.model === "test-embedding";
+    const unauthorized = body.model === "unauthorized-model";
+    const status = routeValid && keyValid && modelValid ? 200 : unauthorized ? 403 : 404;
+    response.writeHead(status, { "Content-Type": "application/json" });
+    response.end(JSON.stringify(status === 200
       ? { data: [{ index: 0, embedding: [1, 0] }] }
-      : { error: "invalid request" }));
+      : { message: "private upstream detail must not leak", request_id: "private-request-id" }));
   });
 });
 server.listen(0, "127.0.0.1", () => {
@@ -122,17 +124,33 @@ EOF
   grep -q "\[PASS\] embedding endpoint accepted model test-embedding" "$tmp/out3" ||
     fail "expected a functional embedding PASS line"
 
-  proj4="$tmp/embedding-wrong-model"; mkdir -p "$proj4/.ai"
+  proj4="$tmp/embedding-unauthorized"; mkdir -p "$proj4/.ai"
   cat > "$proj4/.ai/.env" <<EOF
 CAIRN_MEMORY_EMBEDDING_URL=http://127.0.0.1:$port/v1
-CAIRN_MEMORY_EMBEDDING_MODEL=missing-model
+CAIRN_MEMORY_EMBEDDING_MODEL=unauthorized-model
 CAIRN_LLM_API_KEY=test-key
 EOF
   if ( cd "$proj4" && "$doctor" ) >"$tmp/out4" 2>&1; then
-    fail "doctor should reject a missing embedding model:\n$(cat "$tmp/out4")"
+    fail "doctor should reject an unauthorized embedding request:\n$(cat "$tmp/out4")"
+  fi
+  grep -q "\[FAIL\] embedding authorization rejected (HTTP 403) for model unauthorized-model" "$tmp/out4" ||
+    fail "expected a sanitized embedding authorization diagnostic"
+  ! grep -q "private upstream detail\|private-request-id" "$tmp/out4" ||
+    fail "embedding diagnostic leaked an upstream body or request id"
+
+  proj5="$tmp/embedding-missing-model"; mkdir -p "$proj5/.ai"
+  cat > "$proj5/.ai/.env" <<EOF
+CAIRN_MEMORY_EMBEDDING_URL=http://127.0.0.1:$port/v1
+CAIRN_MEMORY_EMBEDDING_MODEL=missing-model
+CAIRN_LLM_API_KEY=test-key
+EOF
+  if ( cd "$proj5" && "$doctor" ) >"$tmp/out5" 2>&1; then
+    fail "doctor should reject a missing embedding model:\n$(cat "$tmp/out5")"
   fi
-  grep -q "\[FAIL\] embedding request failed for model missing-model" "$tmp/out4" ||
-    fail "expected a functional embedding FAIL line"
+  grep -q "\[FAIL\] embedding endpoint or model not found (HTTP 404) for model missing-model" "$tmp/out5" ||
+    fail "expected a sanitized missing embedding model diagnostic"
+  ! grep -q "private upstream detail\|private-request-id" "$tmp/out5" ||
+    fail "missing-model diagnostic leaked an upstream body or request id"
 fi
 
 # 5. Artifact doctor passes a valid store, repairs only derived state, and