mindforge-cc 10.0.2 → 10.0.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.mindforge/config.json +50 -2
- package/.mindforge/engine/autonomous/cross-iteration-bridge.md +96 -0
- package/.mindforge/engine/cost-tracking/budget-enforcer.md +68 -0
- package/.mindforge/engine/cost-tracking/router.md +58 -0
- package/.mindforge/engine/cost-tracking/token-ledger.md +77 -0
- package/.mindforge/engine/council/council-protocol.md +96 -0
- package/.mindforge/engine/council/council-templates.md +85 -0
- package/.mindforge/engine/council/synthesis-engine.md +71 -0
- package/.mindforge/engine/instincts/capture-engine.md +63 -0
- package/.mindforge/engine/instincts/instinct-schema.md +76 -0
- package/.mindforge/engine/instincts/promotion-engine.md +77 -0
- package/.mindforge/engine/skills/composition.md +83 -0
- package/.mindforge/engine/skills/loader.md +16 -0
- package/.mindforge/personas/cost-optimizer.md +71 -0
- package/.mindforge/personas/council-architect.md +66 -0
- package/.mindforge/personas/council-critic.md +67 -0
- package/.mindforge/personas/council-pragmatist.md +71 -0
- package/.mindforge/personas/council-skeptic.md +73 -0
- package/.mindforge/personas/doc-auditor.md +84 -0
- package/.mindforge/personas/instinct-curator.md +83 -0
- package/.mindforge/personas/multi-model-bridge.md +86 -0
- package/.mindforge/personas/swarm-templates.json +28 -1
- package/.mindforge/personas/threat-modeler.md +82 -0
- package/.mindforge/skills/agent-introspection-debugging/SKILL.md +88 -0
- package/.mindforge/skills/agent-loops/SKILL.md +84 -0
- package/.mindforge/skills/autonomous-loops/SKILL.md +105 -0
- package/.mindforge/skills/continuous-learning/SKILL.md +84 -0
- package/.mindforge/skills/cost-aware-routing/SKILL.md +83 -0
- package/.mindforge/skills/council/SKILL.md +68 -0
- package/.mindforge/skills/doc-health-audit/SKILL.md +102 -0
- package/.mindforge/skills/multi-llm-consult/SKILL.md +75 -0
- package/.mindforge/skills/threat-modeling/SKILL.md +109 -0
- package/.mindforge/skills/verification-loop/SKILL.md +85 -0
- package/CHANGELOG.md +19 -0
- package/MINDFORGE.md +4 -4
- package/README.md +2 -2
- package/RELEASENOTES.md +66 -0
- package/bin/installer-core.js +1 -1
- package/bin/wizard/theme.js +2 -2
- package/docs/commands-reference.md +18 -1
- package/package.json +1 -1
package/RELEASENOTES.md
CHANGED
|
@@ -1,3 +1,69 @@
|
|
|
1
|
+
# Release Notes — v10.0.3 "Council Awakens"
|
|
2
|
+
|
|
3
|
+
**Release Date**: 2026-05-25
|
|
4
|
+
**Tag**: `v10.0.3`
|
|
5
|
+
**Status**: STABLE
|
|
6
|
+
|
|
7
|
+
## Executive Summary
|
|
8
|
+
|
|
9
|
+
MindForge v10.0.3 "Council Awakens" introduces multi-voice deliberation, self-evolving agent instincts, cost-aware routing, and rigorous verification loops. This release expands the framework's autonomous governance and reasoning capabilities with 10 new skills, 8 new commands, 9 new personas, 3 new swarm templates, and 5 engine subsystems.
|
|
10
|
+
|
|
11
|
+
---
|
|
12
|
+
|
|
13
|
+
## Highlights
|
|
14
|
+
|
|
15
|
+
### Council Framework
|
|
16
|
+
- **4-Voice Decision Council** (`/mindforge:council`) — Spawns Architect, Critic, Pragmatist, and Visionary voices for architectural deliberation with structured consensus synthesis.
|
|
17
|
+
- **Multi-LLM Consult** (`/mindforge:consult`) — Routes questions to parallel models (Claude, Gemini, GPT) and synthesizes agreement/disagreement into actionable recommendations.
|
|
18
|
+
|
|
19
|
+
### Instinct Engine
|
|
20
|
+
- **Learn Instinct** (`/mindforge:learn-instinct`) — Captures emergent behavioral patterns from session traces as lightweight instinct primitives.
|
|
21
|
+
- **Evolve Skills** (`/mindforge:evolve-skills`) — Promotes validated instincts through quality scoring into full reusable skills with automated testing.
|
|
22
|
+
|
|
23
|
+
### Cost-Aware Routing
|
|
24
|
+
- **Cost Report** (`/mindforge:cost-report`) — Per-provider token analytics with budget burn rate, daily/weekly trends, and cost-per-task attribution.
|
|
25
|
+
- **Dynamic model selection** based on task complexity vs. budget constraints.
|
|
26
|
+
|
|
27
|
+
### Verification Loop
|
|
28
|
+
- **6-Phase Verification Gates** (`/mindforge:verify-loop`) — Substance, Existence, Wiring, Integration, Regression, and Acceptance gates executed sequentially with fail-fast semantics.
|
|
29
|
+
|
|
30
|
+
### Threat Modeling
|
|
31
|
+
- **STRIDE/DREAD Analysis** (`/mindforge:threat-model`) — Automated attack surface enumeration with severity scoring and mitigation recommendations.
|
|
32
|
+
|
|
33
|
+
### Agent Introspection
|
|
34
|
+
- **Introspect** (`/mindforge:introspect`) — Structured trace analysis for debugging reasoning failures, detecting drift, and identifying optimization opportunities.
|
|
35
|
+
|
|
36
|
+
---
|
|
37
|
+
|
|
38
|
+
## Counts Update
|
|
39
|
+
|
|
40
|
+
| Category | Previous (v10.0.2) | New (v10.0.3) | Delta |
|
|
41
|
+
|----------|-------------------|---------------|-------|
|
|
42
|
+
| Core Skills | 10 | 20 | +10 |
|
|
43
|
+
| Commands | 63 | 71 | +8 |
|
|
44
|
+
| Personas | 108 | 117 | +9 |
|
|
45
|
+
| Swarm Templates | 18 | 21 | +3 |
|
|
46
|
+
| Engine Subsystems | — | +5 | +5 |
|
|
47
|
+
|
|
48
|
+
---
|
|
49
|
+
|
|
50
|
+
## What's Next
|
|
51
|
+
|
|
52
|
+
**v10.1.0** (planned):
|
|
53
|
+
- Coverage target raised to 60% with per-module enforcement
|
|
54
|
+
- OpenTelemetry export for NexusTracer spans (Jaeger/Zipkin compatible)
|
|
55
|
+
- Rate limiting on dashboard API endpoints
|
|
56
|
+
|
|
57
|
+
---
|
|
58
|
+
|
|
59
|
+
## Previous Releases
|
|
60
|
+
|
|
61
|
+
- [v10.0.1 — Bedrock Fortified](https://github.com/sairam0424/MindForge/releases/tag/v10.0.1)
|
|
62
|
+
- [v9.0.0 — Bedrock Meridian](https://github.com/sairam0424/MindForge/releases/tag/v9.0.0)
|
|
63
|
+
|
|
64
|
+
---
|
|
65
|
+
---
|
|
66
|
+
|
|
1
67
|
# Release Notes — v10.0.1 "Bedrock Fortified"
|
|
2
68
|
|
|
3
69
|
**Release Date**: 2026-05-21
|
package/bin/installer-core.js
CHANGED
|
@@ -748,7 +748,7 @@ function collectManifestStats() {
|
|
|
748
748
|
}
|
|
749
749
|
} catch (e) {
|
|
750
750
|
// Fallback to default values if counting fails
|
|
751
|
-
return { personas:
|
|
751
|
+
return { personas: 117, skills: 20, governance: 4, integrations: 6, actions: 71, docs: 12, templates: 21 };
|
|
752
752
|
}
|
|
753
753
|
|
|
754
754
|
return stats;
|
package/bin/wizard/theme.js
CHANGED
|
@@ -94,7 +94,7 @@ const Theme = {
|
|
|
94
94
|
* Success Banner ( V2 Architectural Style)
|
|
95
95
|
*/
|
|
96
96
|
printSuccessV2(runtime, scope, stats = {}) {
|
|
97
|
-
const { personas =
|
|
97
|
+
const { personas = 117, skills = 20, governance = 4, integrations = 7 } = stats;
|
|
98
98
|
const boxWidth = 72;
|
|
99
99
|
|
|
100
100
|
console.log(`\n ${this.colors.green('MINDFORGE is ready! ')} ${this.colors.dim('─'.repeat(boxWidth - 20))}╮`);
|
|
@@ -125,7 +125,7 @@ const Theme = {
|
|
|
125
125
|
* Print Manifest (Hardened for V2)
|
|
126
126
|
*/
|
|
127
127
|
printManifest(stats = {}) {
|
|
128
|
-
const { personas =
|
|
128
|
+
const { personas = 117, skills = 20, governance = 4, integrations = 7, actions = 71, docs = 12, templates = 21 } = stats;
|
|
129
129
|
|
|
130
130
|
console.log(` ${this.colors.bold('PAYLOAD MANIFEST')}`);
|
|
131
131
|
console.log(` ${this.colors.dim('┌' + '─'.repeat(74) + '┐')}`);
|
|
@@ -142,7 +142,24 @@ High-fidelity behaviors ported from the Superpowers library and hardened for pro
|
|
|
142
142
|
|
|
143
143
|
---
|
|
144
144
|
|
|
145
|
-
## 11.
|
|
145
|
+
## 11. Council, Verification & Intelligence (v10.0.3)
|
|
146
|
+
|
|
147
|
+
New commands introduced in the "Council Awakens" release for multi-voice decision-making, verification rigor, and self-evolving agent behavior.
|
|
148
|
+
|
|
149
|
+
| Command | Description |
|
|
150
|
+
| :--- | :--- |
|
|
151
|
+
| `/mindforge:council` | Invoke 4-voice decision council for architectural and strategic deliberation. |
|
|
152
|
+
| `/mindforge:consult` | Multi-LLM second opinion — routes a question to parallel models and synthesizes consensus. |
|
|
153
|
+
| `/mindforge:verify-loop` | Run 6-phase verification gates (substance, existence, wiring, integration, regression, acceptance). |
|
|
154
|
+
| `/mindforge:introspect` | Debug agent behavior and diagnose reasoning failures with structured trace analysis. |
|
|
155
|
+
| `/mindforge:cost-report` | Token usage analytics with per-provider breakdown and budget tracking. |
|
|
156
|
+
| `/mindforge:threat-model` | STRIDE/DREAD threat modeling with automated attack surface enumeration. |
|
|
157
|
+
| `/mindforge:learn-instinct` | Capture learned behavior patterns as instincts for future session priming. |
|
|
158
|
+
| `/mindforge:evolve-skills` | Promote validated instincts into full reusable skills with quality scoring. |
|
|
159
|
+
|
|
160
|
+
---
|
|
161
|
+
|
|
162
|
+
## 12. CLI Commands (`mindforge` binary)
|
|
146
163
|
|
|
147
164
|
After installation, the `mindforge` binary provides direct access to runtime operations. Use `--verbose` (`-v`) on any command for detailed output.
|
|
148
165
|
|
package/package.json
CHANGED