bump(databases-on-aws, deploy-on-aws): f16aaf2a → 187edde

Both plugins in awslabs/agent-plugins had their subpaths edited in commit
187edde (after the morning bump cron pinned them to f16aaf2a), so they fell
behind again on merge. Manual catch-up bump to current monorepo HEAD.

- databases-on-aws: 4 files changed under plugins/databases-on-aws/ (v1.1.0)
- deploy-on-aws:    7 files changed under plugins/deploy-on-aws/ (v1.2.0)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

.claude-plugin/marketplace.json

@@ -120,7 +120,7 @@
         "url": "https://github.com/awslabs/agent-plugins.git",
         "path": "plugins/amazon-location-service",
         "ref": "main",
-        "sha": "f16aaf2a4ec7d59963c4fdf91e7358bd485e992e"
+        "sha": "9d46cc0a092c0a8c01a5bd06a4349985cc6c8f08"
       },
       "homepage": "https://github.com/awslabs/agent-plugins"
     },
@@ -245,7 +245,7 @@
         "url": "https://github.com/auth0/agent-skills.git",
         "path": "plugins/auth0",
         "ref": "main",
-        "sha": "a4fdd4d0605239aa7be75a57371f4e7e71efc40c"
+        "sha": "c38453f6a99bbfeaf73b5be81db987ec6af982da"
       },
       "homepage": "https://auth0.com/docs/quickstart/agent-skills"
     },
@@ -339,6 +339,22 @@
       },
       "homepage": "https://github.com/awslabs/agent-plugins"
     },
+    {
+      "name": "aws-startup-advisor",
+      "description": "Skills for startups building on AWS: an Activate knowledge base (FAQ, credits, programs, partner offers, sample architectures, 277+ learn articles), a prompt library of 29+ copy-paste prompts plus installable agents, and an interactive discovery workflow that scaffolds an AWS architecture.",
+      "author": {
+        "name": "Amazon Web Services"
+      },
+      "category": "development",
+      "source": {
+        "source": "git-subdir",
+        "url": "https://github.com/awslabs/startups.git",
+        "path": "advisor/plugins/aws-startup-advisor",
+        "ref": "main",
+        "sha": "23a4f5eaf74a0f0fcf86f8a05bd36618a3f5faae"
+      },
+      "homepage": "https://github.com/awslabs/startups"
+    },
     {
       "name": "azure",
       "description": "Transform Claude into an Azure expert. This plugin integrates the Azure MCP server and specialized Azure skills to move beyond generic advice. It enables Claude to perform real-world tasks: listing resources, validating deployments, diagnosing infrastructure issues, and optimizing costs across 50+ Azure services.",
@@ -346,7 +362,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/microsoft/azure-skills.git",
-        "sha": "d3440b8a4f138585a512ecd4e0c54ede13ab1cc2"
+        "sha": "7cb89c221ecc9eccb71580aaff3695408cdeef2b"
       },
       "homepage": "https://github.com/microsoft/azure-skills"
     },
@@ -474,7 +490,7 @@
         "url": "https://github.com/carta/plugins.git",
         "path": "plugins/carta-investors",
         "ref": "main",
-        "sha": "f512df80dd54fd0a607ece2f18a9d226ca705019"
+        "sha": "e66d331cd8e669ee121c96ee35b0c91acd828970"
       },
       "homepage": "https://carta.com"
     },
@@ -490,7 +506,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/cap-js/mcp-server.git",
-        "sha": "9658cea90c782a6ab007ac16278c90fa4feca0ed"
+        "sha": "92dc99f5ba0c56957ed5d390484693a69ebd1206"
       },
       "homepage": "https://cap.cloud.sap/"
     },
@@ -716,7 +732,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/CodSpeedHQ/codspeed.git",
-        "sha": "f79d57d207f039e44a31a976564715f7731e71b6"
+        "sha": "407dd3c930b8dc5e5655a2d91a65d88f01829955"
       },
       "homepage": "https://codspeed.io"
     },
@@ -753,7 +769,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/get-convex/convex-backend-skill.git",
-        "sha": "002f9c834cdb834ddef1e4867d87cb6e80f0acba"
+        "sha": "ece93250d560f0ce32a24223dea92b33050b2a66"
       },
       "homepage": "https://github.com/get-convex/convex-backend-skill",
       "keywords": [
@@ -784,7 +800,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/CrowdStrike/foundry-skills.git",
-        "sha": "2908d2f470f14f58d378b26c2af58825fa8d0149"
+        "sha": "fb25d60ecdbc0129071802dad210a65168ca55a9"
       },
       "homepage": "https://github.com/CrowdStrike/foundry-skills"
     },
@@ -878,7 +894,7 @@
         "url": "https://github.com/awslabs/agent-plugins.git",
         "path": "plugins/databases-on-aws",
         "ref": "main",
-        "sha": "f16aaf2a4ec7d59963c4fdf91e7358bd485e992e"
+        "sha": "187edde6e122116b43211049195627a5069bda80"
       },
       "homepage": "https://github.com/awslabs/agent-plugins"
     },
@@ -946,7 +962,7 @@
         "url": "https://github.com/awslabs/agent-plugins.git",
         "path": "plugins/deploy-on-aws",
         "ref": "main",
-        "sha": "f16aaf2a4ec7d59963c4fdf91e7358bd485e992e"
+        "sha": "187edde6e122116b43211049195627a5069bda80"
       },
       "homepage": "https://github.com/awslabs/agent-plugins"
     },
@@ -1114,7 +1130,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/firecrawl/firecrawl-claude-plugin.git",
-        "sha": "81178096fa7ae860285923948b8ba13d03c7fa0c"
+        "sha": "e71cec486062680f0c8f8823afcb3558ad81ce60"
       },
       "homepage": "https://github.com/firecrawl/firecrawl-claude-plugin.git"
     },
@@ -1217,7 +1233,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/huggingface/skills.git",
-        "sha": "49abf82b2ef2ff2bcc6ca072aac0fe8627390a1d"
+        "sha": "df627be1837523c91ac6df472e3dc543d3107bd9"
       },
       "homepage": "https://github.com/huggingface/skills.git"
     },
@@ -1231,7 +1247,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/hunter-io/claude-plugin.git",
-        "sha": "4eb5fbbcb75e0c4c4f2c0d8aa02756165fdde629"
+        "sha": "9b6146520c48f9dcc6092f106e5c1a5762ca3e7a"
       },
       "homepage": "https://hunter.io"
     },
@@ -1245,7 +1261,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/heygen-com/hyperframes.git",
-        "sha": "f8abff2e1c0958b61419b4d2ad91ceca7a7ea110"
+        "sha": "bc3701f5905c5ba7c8cf03c3bbe3a49162d2b1f1"
       },
       "homepage": "https://hyperframes.heygen.com"
     },
@@ -1653,6 +1669,22 @@
       },
       "homepage": "https://github.com/makenotion/claude-code-notion-plugin"
     },
+    {
+      "name": "nvidia-skills",
+      "description": "NVIDIA agent skills for accelerated-computing workflows — starting with cuOpt vehicle-routing optimization (VRP, TSP, PDP) via the cuOpt Python API.",
+      "author": {
+        "name": "NVIDIA"
+      },
+      "category": "development",
+      "source": {
+        "source": "git-subdir",
+        "url": "https://github.com/NVIDIA/skills.git",
+        "path": "plugins/nvidia-skills",
+        "ref": "main",
+        "sha": "62b685a20ac45285cafd1e22782abbed33172c17"
+      },
+      "homepage": "https://github.com/NVIDIA/skills"
+    },
     {
       "name": "oracle-ai-data-platform-workbench-spark-connectors",
       "description": "Oracle AI Data Platform Workbench Spark connectors for Claude Code. 18 connector skills covering every data source workbench customers commonly need: Oracle Autonomous DB family (ALH/ADW/ATP) via wallet/IAM-DB-Token/API-key, ExaCS, Fusion ERP REST, Fusion BICC, EPM Cloud Planning, Essbase 21c, OCI Streaming (Kafka), OCI Object Storage, Apache Iceberg, plus external systems (PostgreSQL, MySQL/HeatWave, SQL Server, Snowflake, Azure ADLS Gen2, AWS S3, generic REST, custom JDBC, Excel). Live-validated on the workbench `tpcds` cluster (Spark 3.5.0): 17 PASS / 4 ship-as-is out of 21 test rows.",
@@ -1665,7 +1697,7 @@
         "url": "https://github.com/oracle-samples/oracle-aidp-samples.git",
         "path": "ai/claude-code-plugins/oracle-ai-data-platform-workbench-spark-connectors",
         "ref": "main",
-        "sha": "d20075e440af2a209bc3b011a336276336e6d10c"
+        "sha": "6e59f24cd3e8870649e7f9b2e3e106502b43fd5f"
       },
       "homepage": "https://docs.oracle.com/en/cloud/paas/ai-data-platform/index.html"
     },
@@ -1681,7 +1713,7 @@
         "url": "https://github.com/growthxai/output.git",
         "path": "coding_assistants/claude/plugins/outputai",
         "ref": "main",
-        "sha": "5a87ebc13343ce6ebabac0bcc443c1da0bcf2459"
+        "sha": "0eeffece25b6f471c48b705a214471164b8c5946"
       },
       "homepage": "https://output.ai"
     },
@@ -1729,7 +1761,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/gopigment/ai-plugins.git",
-        "sha": "abf36e64750d1323a4cc5fe79161597668231224"
+        "sha": "4bf16c80558416b9d69fa6531af8588fb2fcbe27"
       },
       "homepage": "https://www.pigment.com"
     },
@@ -1985,7 +2017,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/Digital-Process-Tools/claude-remember.git",
-        "sha": "a4ff96f38622f7c4920dc349d59cc980663336f4"
+        "sha": "c2c82ab5fd2f4f5c0cddc9c7d8a749655dec4cb9"
       },
       "homepage": "https://github.com/Digital-Process-Tools/claude-remember"
     },
@@ -2113,7 +2145,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/cap-js/mcp-server.git",
-        "sha": "9658cea90c782a6ab007ac16278c90fa4feca0ed"
+        "sha": "92dc99f5ba0c56957ed5d390484693a69ebd1206"
       },
       "homepage": "https://cap.cloud.sap/"
     },
@@ -2131,7 +2163,7 @@
         "url": "https://github.com/SAP/open-ux-tools.git",
         "path": "packages/fiori-mcp-server",
         "ref": "main",
-        "sha": "c3e0940ef29ac520b4cc5bd6fd71fecede7b3342"
+        "sha": "7432d23a7b5c3bd1c0a01cf76696bf0c417ecd1f"
       },
       "homepage": "https://github.com/SAP/open-ux-tools/tree/main/packages/fiori-mcp-server"
     },
@@ -2170,7 +2202,7 @@
     {
       "name": "security-guidance",
       "description": "Security review for Claude-generated code. Pattern-based warnings on edits, LLM-powered diff review on Stop, and an agentic commit reviewer that catches injection, XSS, SSRF, hardcoded secrets, and 25+ other vulnerability classes.",
-      "version": "2.0.0",
+      "version": "2.0.1",
       "author": {
         "name": "Anthropic",
         "email": "support@anthropic.com"
@@ -2198,7 +2230,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/getsentry/sentry-for-claude.git",
-        "sha": "849303a8411c242d250885ffe714235a3bc2f5fe"
+        "sha": "d6123be331e2224b037e1ffefd27c806e7566dcf"
       },
       "homepage": "https://github.com/getsentry/sentry-for-claude/tree/main"
     },
@@ -2214,7 +2246,7 @@
         "url": "https://github.com/getsentry/cli.git",
         "path": "plugins/sentry-cli",
         "ref": "main",
-        "sha": "213c156a47fb2952287938f91e94b5073402530d"
+        "sha": "db90767935558db16c45036f89e68edaa1dde106"
       },
       "homepage": "https://sentry.io"
     },
@@ -2411,7 +2443,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/obra/superpowers.git",
-        "sha": "6fd4507659784c351abbd2bc264c7162cfd386dc"
+        "sha": "f2cbfbefebbfef77321e4c9abc9e949826bea9d7"
       },
       "homepage": "https://github.com/obra/superpowers.git"
     },
@@ -2445,7 +2477,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/JetBrains/teamcity-cli.git",
-        "sha": "533c8cb20928be912188fd8ae40fcba24cc720ca"
+        "sha": "9436b94b228579ba952aba809357776c3db9ce1a"
       },
       "homepage": "https://www.jetbrains.com/teamcity/"
     },
@@ -2620,7 +2652,7 @@
       "source": {
         "source": "url",
         "url": "https://github.com/wix/skills.git",
-        "sha": "357f8d5b7cd81455e6a57cc64dba83985131d78f"
+        "sha": "c5b343f2dadba06da91ee6de07272161fb68d40d"
       },
       "homepage": "https://dev.wix.com/docs/wix-cli/guides/development/about-wix-skills"
     },

.github/workflows/bump-plugin-shas.yml

@@ -2,25 +2,24 @@ name: Bump Plugin SHAs
 
 # Nightly sweep: for each external entry whose upstream HEAD has moved past
 # its pinned SHA, validate at the new SHA with `claude plugin validate`
-# inline, then open one PR with all passing bumps. Each run force-resets the
-# bump/plugin-shas branch, so a previous night's unmerged PR is replaced (and
-# its review state discarded) — review and merge same-day to avoid churn.
+# inline, then open one PR per bumped plugin on branch `bump/<slug>`.
+# Failing entries stay isolated in their own PR; passing bumps merge
+# independently.
 #
 # Bot-free — uses the default GITHUB_TOKEN. PRs opened with GITHUB_TOKEN don't
-# trigger on:pull_request workflows, so the policy scan (`Scan Plugins`, a
-# required status check on main) would never run and the bump PR could never
-# merge. workflow_dispatch is exempt from that recursion guard, so we dispatch
-# the scan ourselves on the bump branch after the PR is opened. The check run
-# lands on the branch HEAD — the same SHA as the PR head — and satisfies the
-# required check.
+# trigger on:pull_request workflows, so the required status checks on main
+# (`scan` from Scan Plugins, `check` from Check MCP URLs, `validate` from
+# Validate Plugins) would never run and the bump PR could never merge.
+# workflow_dispatch is exempt from that recursion guard, so we dispatch all
+# three ourselves against each per-entry bump branch after its PR is opened.
+# Each check run lands on the branch HEAD — the same SHA as the PR head — and
+# satisfies the corresponding required check. (Each of those workflows runs
+# its job unconditionally on workflow_dispatch, so a dispatch always reports.)
 #
-# max-bumps is set above the external-entry count so a single run can clear
-# any backlog. The cost-control mechanisms are downstream:
-#   - scan-plugins.yml caches verdicts by (plugin, sha) so an unchanged SHA
-#     is never re-scanned across nightly force-resets.
-#   - revert-failed-bumps.yml drops policy-failing entries from the bump PR
-#     so one bad upstream can't block the rest.
-# See those files for details.
+# max-bumps caps the per-night work for cost control. Per-entry scans are
+# more expensive than a single batched scan, so the cap is conservative.
+# The composite action skips entries that already have an open bump PR, so
+# re-dispatches don't pile up duplicate work.
 
 on:
   schedule:
@@ -30,12 +29,12 @@ on:
       max_bumps:
         description: Cap on plugins bumped this run
         required: false
-        default: '130'
+        default: '30'
 
 permissions:
   contents: write
   pull-requests: write
-  actions: write  # gh workflow run scan-plugins.yml on the bump branch
+  actions: write  # gh workflow run {scan-plugins,check-mcp-urls,validate-plugins}.yml per bump branch
 
 concurrency:
   group: bump-plugin-shas
@@ -43,8 +42,8 @@ concurrency:
 jobs:
   bump:
     runs-on: ubuntu-latest
-    # Per-bump cost is ~2s (ls-remote + shallow clone + validate); 130 entries
-    # is ~5 min. The 60 min ceiling absorbs slow upstreams without letting a
+    # Per-bump cost is ~2s (ls-remote + shallow clone + validate); 30 entries
+    # is ~1-2 min. The 60 min ceiling absorbs slow upstreams without letting a
     # pathological run consume the default 360 min budget.
     timeout-minutes: 60
     steps:
@@ -52,18 +51,44 @@ jobs:
 
       # createCommitOnBranch-based bump so commits are signed by GitHub and
       # satisfy the org-level required_signatures ruleset on main.
-      - uses: anthropics/claude-plugins-community/.github/actions/bump-plugin-shas@c41c6911de0afffd2bc5cd8b21fb1e06444ee13b
+      - uses: anthropics/claude-plugins-community/.github/actions/bump-plugin-shas@e2019b2a01f11aa1484c53540b1cfab5eebbc299
         id: bump
         with:
           marketplace-path: .claude-plugin/marketplace.json
-          max-bumps: ${{ inputs.max_bumps || '130' }}
+          max-bumps: ${{ inputs.max_bumps || '30' }}
+          pr-mode: per-entry
           claude-cli-version: latest
 
-      # `bump/plugin-shas` is the action's default `pr-branch`. The scan diffs
-      # the branch against origin/main (the action's base-ref fallback when
-      # there's no pull_request event) and scans only the bumped entries.
-      - name: Dispatch policy scan on bump branch
-        if: steps.bump.outputs.pr-url != ''
+      # Per-entry fan-out: dispatch the three required checks against each bump
+      # branch. `pr-urls` is a JSON array of {name, old_sha, new_sha, branch,
+      # pr_url} entries emitted by the composite action when pr-mode is
+      # per-entry. All three (scan / check / validate) are required on main and
+      # none fire on the GITHUB_TOKEN-opened PR, so each must be dispatched.
+      # A single failed dispatch (transient API error / rate limit) must not
+      # strand the remaining branches, so we attempt every dispatch, then fail
+      # the step if any failed: a missing required check would otherwise leave
+      # its bump PR silently blocked behind a green run, and the composite
+      # action skips slugs with an open PR so it would never be retried.
+      - name: Dispatch required checks per per-entry PR
+        if: steps.bump.outputs.pr-urls != '' && steps.bump.outputs.pr-urls != '[]'
         env:
           GH_TOKEN: ${{ github.token }}
-        run: gh workflow run scan-plugins.yml --ref bump/plugin-shas
+          PR_URLS: ${{ steps.bump.outputs.pr-urls }}
+        run: |
+          set -euo pipefail
+          dispatch_failures="$(mktemp)"
+          jq -c '.[]' <<<"$PR_URLS" | while read -r entry; do
+            branch=$(jq -r '.branch' <<<"$entry")
+            name=$(jq -r '.name' <<<"$entry")
+            for wf in scan-plugins check-mcp-urls validate-plugins; do
+              echo "Dispatching ${wf}.yml against $branch ($name)"
+              if ! gh workflow run "${wf}.yml" --ref "$branch"; then
+                echo "::error::Failed to dispatch ${wf}.yml against $branch ($name) — required check will be missing; re-dispatch with: gh workflow run ${wf}.yml --ref $branch"
+                echo "${wf} ${branch}" >> "$dispatch_failures"
+              fi
+            done
+          done
+          if [ -s "$dispatch_failures" ]; then
+            echo "::error::$(wc -l < "$dispatch_failures" | tr -d ' ') required-check dispatch(es) failed; the affected bump PR(s) are blocked until re-dispatched (see annotations above)."
+            exit 1
+          fi

.github/workflows/validate-plugins.yml

@@ -12,6 +12,14 @@ on:
     branches: [main]
     paths:
       - '.claude-plugin/**'
+  # `validate` is a required status check on main. Bump PRs are opened with
+  # GITHUB_TOKEN, which doesn't fire on:pull_request (recursion guard), so the
+  # path-filtered trigger above never reports on them and the PR would be
+  # blocked forever. The bump workflow dispatches this against each per-entry
+  # bump branch instead; the check run lands on the branch HEAD (= PR head)
+  # and satisfies the required check. The validate job runs unconditionally,
+  # so a dispatch always reports.
+  workflow_dispatch:
 
 permissions:
   contents: read

plugins/security-guidance/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "security-guidance",
-  "version": "2.0.0",
+  "version": "2.0.1",
   "description": "Security review for Claude-generated code. Pattern-based warnings on edits, LLM-powered diff review on Stop, and an agentic commit reviewer that catches injection, XSS, SSRF, hardcoded secrets, and 25+ other vulnerability classes.",
   "author": {
     "name": "David Dworken",

plugins/security-guidance/hooks/_base.py

@@ -116,7 +116,18 @@ _PV = _read_plugin_version_int()
 # Emitted via _usage_metrics() into the existing emit_metrics() channel so
 # hook metrics rows carry per-invocation token/cost totals
 # alongside the existing skip_reason / vulns_found fields.
-_USAGE = {"in": 0, "out": 0, "cr": 0, "cw": 0, "cost": 0.0, "n": 0}
+_USAGE = {
+    "in": 0, "out": 0, "cr": 0, "cw": 0, "cost": 0.0, "n": 0,
+    # HTTP error visibility (#2098 visibility gap — see emit comment in
+    # _usage_metrics). Without this, API failures from `_call_claude` left
+    # zero fingerprint in telemetry: the call returns None, the caller's
+    # emit_metrics carries no api_calls field, and the failure is
+    # indistinguishable from "no review needed". The deprecation outage
+    # that broke every commit-review LLM call was invisible until users
+    # reported it manually.
+    "http_err_last": 0,    # most recent HTTP error code this invocation
+    "http_err_count": 0,   # total HTTP errors (4xx + 5xx + network)
+}
 _USAGE_LOCK = threading.Lock()
 
 # $/Mtok (input, output). Used only for the raw-HTTP path; the SDK path
@@ -166,19 +177,55 @@ def _record_usage(usage, model, cost_usd=None):
         _USAGE["n"] += 1
 
 
+def _record_http_error(status):
+    """Record an HTTP error from an LLM API call. `status` is the HTTP
+    status code (integer 400–599) or -1 for network/timeout errors. Stored
+    in `_USAGE["http_err_last"]` (most recent) and counted in
+    `_USAGE["http_err_count"]`. Snapshot via `_usage_metrics()` so every
+    subsequent `emit_metrics` includes the failure fingerprint.
+
+    Background: without this, the most recent example was the #2098
+    deprecation 400. Every hook fire's LLM call returned HTTP 400; the
+    plugin caught it and returned None; the emit_metrics carried no
+    api_calls field; aggregate dashboards looked normal. The failure
+    only became visible when a user manually reported errors out of
+    their debug log. With this field, a category-of-failure spike (4xx,
+    5xx, or -1 network) is queryable from BQ in real time.
+    """
+    try:
+        s = int(status)
+    except (TypeError, ValueError):
+        return
+    with _USAGE_LOCK:
+        _USAGE["http_err_last"] = s
+        _USAGE["http_err_count"] += 1
+
+
 def _usage_metrics():
     """Snapshot the accumulator as metric keys. Returns {} when no API calls
-    were made so skip-path emits don't burn key budget. cost_usd rounded to
-    1e-6 to keep the float finite/short for the zod schema."""
-    with _USAGE_LOCK:
-        if _USAGE["n"] == 0:
-            return {}
-        return {
-            "tok_in": _USAGE["in"],
-            "tok_out": _USAGE["out"],
-            "tok_cache_r": _USAGE["cr"],
-            "tok_cache_w": _USAGE["cw"],
-            "cost_usd": round(_USAGE["cost"], 6),
-            "api_calls": _USAGE["n"],
-        }
+    AND no HTTP errors were made so skip-path emits don't burn key budget.
+    cost_usd rounded to 1e-6 to keep the float finite/short for the zod
+    schema.
+
+    HTTP errors (`http_err_last`, `http_err_count`) emitted ONLY when
+    `http_err_count > 0` so successful calls don't pad every metrics row
+    with two zero fields.
+    """
+    with _USAGE_LOCK:
+        if _USAGE["n"] == 0 and _USAGE["http_err_count"] == 0:
+            return {}
+        out = {}
+        if _USAGE["n"] > 0:
+            out.update({
+                "tok_in": _USAGE["in"],
+                "tok_out": _USAGE["out"],
+                "tok_cache_r": _USAGE["cr"],
+                "tok_cache_w": _USAGE["cw"],
+                "cost_usd": round(_USAGE["cost"], 6),
+                "api_calls": _USAGE["n"],
+            })
+        if _USAGE["http_err_count"] > 0:
+            out["http_err_last"] = _USAGE["http_err_last"]
+            out["http_err_count"] = _USAGE["http_err_count"]
+        return out
 

plugins/security-guidance/hooks/ensure_agent_sdk.py

@@ -42,6 +42,71 @@ HOOK_PY_INCOMPATIBLE = 6  # hook interpreter is <3.10 — SDK syntax can't load
                           # here no matter how the venv was built. See #2071.
 
 
+# Phase + err-kind integer encoding for sdk_bootstrap_phase / sdk_bootstrap_err.
+#
+# Earlier versions emitted these as STRINGS (e.g. "pip", "dns_fail"). CC's
+# plugin-metrics pipeline silently drops plugin-emitted string values —
+# only `bool|finite-number` plugin metrics reach BigQuery. (CC-core
+# metrics like `subscription_type` are exempt because they're injected
+# downstream of plugin validation.) Confirmed empirically: 185K
+# BUILD_FAILED rows in BQ had `sdk_bootstrap_phase`/`sdk_bootstrap_err`
+# = NULL despite the Python code emitting them. This left ~28K
+# BUILD_FAILED sessions/day with no diagnostic split — flying blind on
+# the real failure modes (pip-no-match vs dns-fail vs ssl-verify etc.).
+#
+# Fix: encode as small integers per the maps below. Values are
+# APPEND-ONLY for telemetry stability. Reserve 99 as the "unknown /
+# uncategorized" bucket so an unmapped err_kind (e.g., a new exception
+# type) still emits a non-zero signal.
+SDK_BOOTSTRAP_PHASE_CODES = {
+    "pre":  1,  # pre-venv (state_dir.mkdir, sentinel open)
+    "venv": 2,  # python -m venv --clear
+    "pip":  3,  # pip install
+    "main": 4,  # uncaught exception above main()
+}
+SDK_BOOTSTRAP_ERR_CODES = {
+    "pip_no_match":       1,
+    "dns_fail":           2,
+    "conn_refused":       3,
+    "ssl_verify":         4,
+    "perm_denied":        5,
+    "no_pip":             6,
+    "disk_full":          7,
+    "proxy_auth":         8,
+    "stderr_timeout":     9,  # pip stderr containing "timeout"/"timed out"
+    "subprocess_timeout": 10, # subprocess.TimeoutExpired (>120s)
+    # 11–98 reserved for future categories; APPEND-ONLY.
+    # 99 catches everything else (including "exc:<TypeName>" and "other:<tail>"
+    # — the original string is debug-loggable but the integer is what makes
+    # it to telemetry).
+    "_uncategorized":     99,
+}
+
+
+def _encode_phase(s):
+    """Map err_phase string to its telemetry integer code, or 0 if unset.
+    Empty/None → 0 lets `if encoded:` cleanly skip emission. Per
+    SDK_BOOTSTRAP_PHASE_CODES, valid codes are 1-4."""
+    return SDK_BOOTSTRAP_PHASE_CODES.get((s or "").strip(), 0)
+
+
+def _encode_err_kind(s):
+    """Map err_kind string to its telemetry integer code, or 0 if unset.
+    Direct hits use the static map; "exc:<X>" and "other:<tail>" both
+    collapse to _uncategorized (99) — the raw string survives in debug
+    logs, only the integer reaches BQ."""
+    s = (s or "").strip()
+    if not s:
+        return 0
+    if s in SDK_BOOTSTRAP_ERR_CODES:
+        return SDK_BOOTSTRAP_ERR_CODES[s]
+    # Prefix matches for the catch-all categories
+    if s.startswith("exc:") or s.startswith("other:") or s == "other":
+        return SDK_BOOTSTRAP_ERR_CODES["_uncategorized"]
+    # Unknown string — still emit as uncategorized rather than dropping
+    return SDK_BOOTSTRAP_ERR_CODES["_uncategorized"]
+
+
 def _sdk_on_syspath() -> bool:
     # find_spec is ~10ms; actually importing the SDK pulls in
     # transitive deps and costs ~800ms — too heavy for a
@@ -288,21 +353,25 @@ if __name__ == "__main__":
     # and takes the FIRST non-{"async":...} JSON line as the hook response;
     # its `metrics` key is forwarded to the hook metrics event on the
     # next attachments pass. Must be a single line — the registry splits on
-    # \n and json-parses each independently. Values must be bool|number OR
-    # short strings (CC accepts string metric values if they're not
-    # null). Stay inside the 10-key emit cap.
+    # \n and json-parses each independently.
+    #
+    # IMPORTANT — values must be bool|finite-number. The validation comment
+    # has historically said "or short strings" but that was wrong: CC's
+    # plugin-metrics pipeline silently drops plugin-emitted string values.
+    # Stay inside the 10-key emit cap.
     metrics: dict[str, object] = {
         "sdk_bootstrap": outcome,
         "sdk_bootstrap_ms": round((time.perf_counter() - t0) * 1000),
     }
     if err_kind:
-        # Truncate defensively; categorized values are <40 chars but the
-        # `other:<tail>` mode could be longer. err_phase may be empty for
-        # pre-venv failures (state_dir.mkdir perm-denied, sentinel O_EXCL
-        # raising a non-FileExistsError OSError) — emit as "pre" so the
-        # err_kind isn't silently dropped.
-        metrics["sdk_bootstrap_phase"] = (err_phase or "pre")[:16]
-        metrics["sdk_bootstrap_err"] = err_kind[:96]
+        # Encode phase + err_kind as integer codes (see
+        # SDK_BOOTSTRAP_PHASE_CODES / SDK_BOOTSTRAP_ERR_CODES). Earlier
+        # versions emitted these as strings and CC dropped them — restoring
+        # the diagnostic split that 28K BUILD_FAILED/day need to triage by
+        # root cause. err_phase defaults to "pre" when empty (pre-venv
+        # failure path, e.g. state_dir.mkdir perm-denied).
+        metrics["sdk_bootstrap_phase"] = _encode_phase(err_phase or "pre")
+        metrics["sdk_bootstrap_err"] = _encode_err_kind(err_kind)
     pv = _plugin_version_int()
     if pv:
         metrics["pv"] = pv

plugins/security-guidance/hooks/llm.py

@@ -27,7 +27,7 @@ from typing import Optional, Tuple, Dict, Any, List
 
 import extensibility
 import review_api
-from _base import debug_log, _record_usage, _PV, PROVENANCE_TAG, state_dir as _resolve_state_dir  # noqa: F401
+from _base import debug_log, _record_usage, _record_http_error, _PV, PROVENANCE_TAG, state_dir as _resolve_state_dir  # noqa: F401
 from session_state import with_locked_state
 
 
@@ -368,6 +368,7 @@ def _call_claude_via_sdk(prompt, output_schema, *, max_tokens=16000, model=None)
         except Exception as e:
             debug_log(f"3P sdk-single-turn: SDK unavailable ({e})")
             _last_call_claude_http_error = -1
+            _record_http_error(-1)
             return None
 
     cli_path = os.environ.get("SG_AGENTIC_CLI_PATH") or None
@@ -425,6 +426,7 @@ def _call_claude_via_sdk(prompt, output_schema, *, max_tokens=16000, model=None)
     except _asyncio.TimeoutError:
         debug_log("3P sdk-single-turn: timeout after 60s")
         _last_call_claude_http_error = -1
+        _record_http_error(-1)
         return None
     except Exception as e:
         debug_log(f"3P sdk-single-turn: query failed ({e})")
@@ -433,6 +435,7 @@ def _call_claude_via_sdk(prompt, output_schema, *, max_tokens=16000, model=None)
             for _l in _captured_stderr[:20]:
                 debug_log(f"  | {_l.rstrip()}")
         _last_call_claude_http_error = -1
+        _record_http_error(-1)
         return None
 
 
@@ -542,6 +545,7 @@ def _call_claude(prompt, output_schema, thinking_budget=10000, max_tokens=16000,
                 error_body = e.read().decode("utf-8") if e.fp else ""
                 debug_log(f"API error: {e.code} - {error_body[:200]}")
                 _last_call_claude_http_error = e.code
+                _record_http_error(e.code)
                 return None
         except (urllib.error.URLError, TimeoutError) as e:
             if attempt < 2:
@@ -551,6 +555,7 @@ def _call_claude(prompt, output_schema, thinking_budget=10000, max_tokens=16000,
             else:
                 debug_log(f"Request failed after retries: {e}")
                 _last_call_claude_http_error = -1
+                _record_http_error(-1)
                 return None
 
     if not response_data:
@@ -559,6 +564,7 @@ def _call_claude(prompt, output_schema, thinking_budget=10000, max_tokens=16000,
         # call uses the token; record the 401 so callers don't see error=None.
         if _last_call_claude_http_error is None:
             _last_call_claude_http_error = 401
+            _record_http_error(401)
         return None
 
     # Find the text block (skip thinking blocks)