Skip to content

Commit 235ed8d

Browse files
proffesor-for-testingproffesor-for-testing
committed
chore(release): bump version to v3.11.5
Bumps package.json + package-lock + skills-manifest fleetVersion to 3.11.5. Also fixes .claude-plugin/plugin.json and assets/skills/skills-manifest.json fleetVersion, which had drifted stale at 3.11.0 across the last 4 releases (found via full-repo version-string audit).
1 parent a454889 commit 235ed8d

8 files changed

Lines changed: 129 additions & 6 deletions

File tree

‎.claude-plugin/plugin.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
{
22
"name": "agentic-qe",
3-
"version": "3.11.0",
3+
"version": "3.11.5",
44
"description": "Agentic Quality Engineering — AI-powered QE platform with 60 specialized agents, 75+ skills, sublinear coverage analysis, ReasoningBank pattern learning, and deep MCP integration for Claude Code and 11 coding agent platforms",
55
"author": {
66
"name": "Agentic QE Team",

‎.claude/skills/skills-manifest.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -939,7 +939,7 @@
939939
},
940940
"metadata": {
941941
"generatedBy": "Agentic QE Fleet",
942-
"fleetVersion": "3.11.4",
942+
"fleetVersion": "3.11.5",
943943
"manifestVersion": "1.4.0",
944944
"lastUpdated": "2026-04-13T00:00:00.000Z",
945945
"contributors": [

‎CHANGELOG.md‎

Lines changed: 58 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -5,6 +5,64 @@ All notable changes to the Agentic QE project will be documented in this file.
55
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
66
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
77

8+
## [3.11.5] - 2026-07-07
9+
10+
A system-integrity sweep of the self-learning loop: dream-cycle insights now
11+
genuinely turn into reusable patterns instead of silently piling up, pattern
12+
usage stats are accurate again (no more double-counting or mid-session
13+
regressions), and the witness-chain audit trail, agent-topology tracking, GOAP
14+
plan execution, and SONA continual learning all had real dormant-code or
15+
data-integrity bugs fixed. See
16+
[SYSTEM-INTEGRITY-AUDIT-2026-07-04](docs/analysis/SYSTEM-INTEGRITY-AUDIT-2026-07-04.md)
17+
and the [remediation plan](docs/plans/SYSTEM-INTEGRITY-REMEDIATION-GOAP-PLAN-2026-07-04.md)
18+
for full findings and verification evidence.
19+
20+
### Fixed
21+
22+
- **Dream-cycle insights now genuinely become reusable patterns** — a gate in
23+
the concept-loading path meant the fix that lets `detectGaps`/
24+
`detectOptimizations`/`detectPatternMerges` see real failure/success data
25+
only ever ran in manual tests, never in real scheduled dream cycles. Also
26+
fixed the mechanism that marks an insight "applied": it previously
27+
incremented a counter on the 3 newest insights on every successful task
28+
regardless of whether anything was actually promoted (proven live in
29+
production, with counts as high as 16) — it now only marks an insight
30+
applied when it's genuinely turned into a new pattern.
31+
- **`aqe learning stats` reported the wrong pattern counts** — Total and "By
32+
Domain" now come from the real pattern store instead of an incomplete
33+
vector-index count that silently excluded any pattern without an embedding.
34+
- **Pattern usage was recorded twice per outcome**, inflating usage counts and
35+
success-rate stats 2x; now written once.
36+
- **Pattern lookups could silently fail after a name collision** ("Pattern not
37+
found") because an internal id could go out of sync between the in-memory
38+
cache and the database; fixed for both pattern-store backends.
39+
- **Learning stats could regress mid-session** — reported totals could drop
40+
below the real historical count right after the first pattern-usage event
41+
in a session; now always reports the correct, non-decreasing total.
42+
- **Witness-chain audit trail** — cryptographic signing now covers all new
43+
entries, and a bug that silently broke chain verification for every entry
44+
after an archival operation is fixed, plus a new CI gate to catch
45+
regressions.
46+
- **Agent-topology tracking (mincut)** now reflects real agent activity
47+
instead of a constant placeholder reading, removing a false "critical"
48+
routing state that could occur on an empty graph.
49+
- **GOAP plan execution** now dispatches real domain-API calls for the
50+
majority of actions instead of always simulating (mocking) results.
51+
- **SONA continual-learning cold start** — the self-learning weight-update
52+
mechanism no longer needs 100 requests to accumulate within a single
53+
process lifetime before it can save its first update; progress now
54+
persists across restarts.
55+
- Assorted smaller integrity fixes: pattern-null (failure record) capture
56+
wasn't wired into the production learning path; a stale/expired daemon
57+
process could look "healthy"; a TTL bug caused several cache entries to
58+
live 1000x longer than intended.
59+
60+
### Changed
61+
62+
- Cleaned up dead data left over from now-fixed bugs (degenerate agent-topology
63+
snapshots, orphaned GOAP test-fixture rows, a corrupted insight-tracking
64+
counter) — internal only, no user-facing behavior change.
65+
866
## [3.11.4] - 2026-07-01
967

1068
Better free local-model support, a tamper-evident guard for AQE's learning

‎assets/skills/skills-manifest.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -939,7 +939,7 @@
939939
},
940940
"metadata": {
941941
"generatedBy": "Agentic QE Fleet",
942-
"fleetVersion": "3.11.0",
942+
"fleetVersion": "3.11.5",
943943
"manifestVersion": "1.4.0",
944944
"lastUpdated": "2026-04-13T00:00:00.000Z",
945945
"contributors": [

‎docs/releases/README.md‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,7 @@ All Agentic QE release notes organized by version.
44

55
| Version | Date | Highlights |
66
|---------|------|------------|
7+
| [v3.11.5](v3.11.5.md) | 2026-07-07 | System-integrity sweep of the self-learning loop: dream-cycle insights now genuinely become reusable patterns (gate + fake-counter fixes); pattern usage stats fixed (double-write, id-desync lookup failures, mid-session regression); witness-chain signing/archival, agent-topology tracking, GOAP real execution, and SONA cold-start bugs fixed. |
78
| [v3.11.4](v3.11.4.md) | 2026-07-01 | Free-tier local test-gen defaults to code-tuned `qwen3-coder:30b` + pattern-cache skip; unified `AQE_OLLAMA_URL` + adaptive timeouts + capability-aware tool routing; opt-in proof-gate for tamper-evident `memory.db` integrity (ADR-116); self-running dependency-security system (daily audit + weekly in-range PRs, ADR-115); grpc-js HIGH CVE fixed, moderates 29→7. Additive. |
89
| [v3.11.3](v3.11.3.md) | 2026-06-27 | Oracle evals (run + mutate generated tests, ADR-113), durable-first test generation, mutation-score/regenerability gate, conservation guard for CLI/output/skill surfaces (ADR-114), cheaper-model lanes; fixed OpenRouter + Gemini providers. Additive. |
910
| [v3.11.2](v3.11.2.md) | 2026-06-27 | First-class C4 architecture diagrams from your codebase: `aqe code c4` CLI + `qe/code/c4` MCP tool generate Context/Container/Component Mermaid with a deterministic confidence gate, real Knowledge-Graph-derived relationships, and semantic diagram search. Additive. |

‎docs/releases/v3.11.5.md‎

Lines changed: 64 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,64 @@
1+
# v3.11.5 Release Notes
2+
3+
**Release Date:** 2026-07-07
4+
5+
## Highlights
6+
7+
A system-integrity sweep of the self-learning loop: **dream-cycle insights now
8+
genuinely turn into reusable patterns** instead of silently piling up (a gate
9+
bug meant the fix only ever ran in manual tests, never in real scheduled
10+
cycles), and **pattern usage stats are accurate again** — no more
11+
double-counting, silent lookup failures after name collisions, or mid-session
12+
regressions. The **witness-chain audit trail**, **agent-topology tracking**,
13+
**GOAP plan execution**, and **SONA continual learning** each had real
14+
dormant-code or data-integrity bugs fixed. See
15+
[SYSTEM-INTEGRITY-AUDIT-2026-07-04](../analysis/SYSTEM-INTEGRITY-AUDIT-2026-07-04.md)
16+
and the
17+
[remediation plan](../plans/SYSTEM-INTEGRITY-REMEDIATION-GOAP-PLAN-2026-07-04.md)
18+
for full findings and verification evidence.
19+
20+
## Fixed
21+
22+
- **Dream-cycle insights now genuinely become reusable patterns** — a gate in
23+
the concept-loading path meant the fix that lets `detectGaps`/
24+
`detectOptimizations`/`detectPatternMerges` see real failure/success data
25+
only ever ran in manual tests, never in real scheduled dream cycles. Also
26+
fixed the mechanism that marks an insight "applied": it previously
27+
incremented a counter on the 3 newest insights on every successful task
28+
regardless of whether anything was actually promoted (proven live in
29+
production, with counts as high as 16) — it now only marks an insight
30+
applied when it's genuinely turned into a new pattern.
31+
- **`aqe learning stats` reported the wrong pattern counts** — Total and "By
32+
Domain" now come from the real pattern store instead of an incomplete
33+
vector-index count that silently excluded any pattern without an embedding.
34+
- **Pattern usage was recorded twice per outcome**, inflating usage counts and
35+
success-rate stats 2x; now written once.
36+
- **Pattern lookups could silently fail after a name collision** ("Pattern not
37+
found") because an internal id could go out of sync between the in-memory
38+
cache and the database; fixed for both pattern-store backends.
39+
- **Learning stats could regress mid-session** — reported totals could drop
40+
below the real historical count right after the first pattern-usage event
41+
in a session; now always reports the correct, non-decreasing total.
42+
- **Witness-chain audit trail** — cryptographic signing now covers all new
43+
entries, and a bug that silently broke chain verification for every entry
44+
after an archival operation is fixed, plus a new CI gate to catch
45+
regressions.
46+
- **Agent-topology tracking (mincut)** now reflects real agent activity
47+
instead of a constant placeholder reading, removing a false "critical"
48+
routing state that could occur on an empty graph.
49+
- **GOAP plan execution** now dispatches real domain-API calls for the
50+
majority of actions instead of always simulating (mocking) results.
51+
- **SONA continual-learning cold start** — the self-learning weight-update
52+
mechanism no longer needs 100 requests to accumulate within a single
53+
process lifetime before it can save its first update; progress now
54+
persists across restarts.
55+
- Assorted smaller integrity fixes: pattern-null (failure record) capture
56+
wasn't wired into the production learning path; a stale/expired daemon
57+
process could look "healthy"; a TTL bug caused several cache entries to
58+
live 1000x longer than intended.
59+
60+
## Changed
61+
62+
- Cleaned up dead data left over from now-fixed bugs (degenerate agent-topology
63+
snapshots, orphaned GOAP test-fixture rows, a corrupted insight-tracking
64+
counter) — internal only, no user-facing behavior change.

‎package-lock.json‎

Lines changed: 2 additions & 2 deletions
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

‎package.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
{
22
"name": "agentic-qe",
3-
"version": "3.11.4",
3+
"version": "3.11.5",
44
"description": "Agentic Quality Engineering V3 - Domain-Driven Design Architecture with 13 Bounded Contexts, O(log n) coverage analysis, ReasoningBank learning, 60 specialized QE agents, mathematical Coherence verification, deep Claude Flow integration",
55
"type": "module",
66
"main": "./dist/index.js",

0 commit comments

Comments
 (0)