math-skill 3.3.0 → 3.3.2
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/README.en-US.md +141 -136
- package/README.md +144 -139
- package/SKILL.en.md +1 -1
- package/SKILL.md +1 -1
- package/bin/math-skill.cjs +276 -0
- package/commands/ask.en.md +3 -2
- package/commands/ask.md +3 -2
- package/package.json +13 -35
- package/references/inspiration.en.md +2 -6
- package/references/inspiration.md +3 -7
- package/references/skill-index.en.md +3 -3
- package/references/skill-index.md +3 -3
- package/skills/math-research-activator/SKILL.en.md +1 -1
- package/skills/math-research-activator/SKILL.md +1 -1
- package/skills/math-research-activator/original-texts.en.md +8 -6
- package/skills/math-research-activator/original-texts.md +4 -2
package/README.en-US.md
CHANGED
|
@@ -1,64 +1,111 @@
|
|
|
1
|
-
<
|
|
2
|
-
<a href="README.md">中文</a> | <a href="README.en-US.md">English</a>
|
|
3
|
-
</p>
|
|
1
|
+
<div align="center">
|
|
4
2
|
|
|
5
|
-
# 📐 Math Skill
|
|
3
|
+
# 📐 Math Skill
|
|
4
|
+
|
|
5
|
+
### A Mathematical Research Operating System for AI and Cryptography Innovation
|
|
6
|
+
|
|
7
|
+
<a href="README.md">中文</a> | <a href="README.en-US.md">English</a>
|
|
6
8
|
|
|
7
9
|
[](LICENSE)
|
|
8
10
|
[](https://www.npmjs.com/package/math-skill)
|
|
9
11
|
[](https://www.npmjs.com/package/math-skill)
|
|
10
12
|
|
|
11
|
-
|
|
13
|
+
</div>
|
|
14
|
+
|
|
15
|
+
> Math Skill does not store mathematics. It activates, routes, and translates mathematics for AI and cryptography research.
|
|
16
|
+
|
|
17
|
+
A math reasoning skill for **Claude Code / Codex / Cursor and more** — automatically activates the right mathematical lenses, knowledge anchors, and design prototypes for AI architecture design, math-to-AI transfer, and cryptographic proof review, delivering conditional, falsifiable conclusions with counterexamples.
|
|
12
18
|
|
|
13
|
-
|
|
19
|
+
If this skill inspires you, please consider leaving a Star⭐. Your support keeps the project evolving.
|
|
14
20
|
|
|
15
21
|
---
|
|
16
22
|
|
|
17
|
-
##
|
|
23
|
+
## How is this different from just asking the AI?
|
|
18
24
|
|
|
19
|
-
|
|
25
|
+
### [AI] Case 1: Analysis — "KAN's mathematical foundations and risks"
|
|
20
26
|
|
|
21
|
-
|
|
27
|
+
**User**: "KAN uses learnable B-splines instead of ReLU, claiming better accuracy and interpretability than MLPs. I want to analyze its mathematical foundations and potential risks."
|
|
22
28
|
|
|
23
|
-
|
|
29
|
+
**Without Skill**:
|
|
30
|
+
> KAN is based on the Kolmogorov-Arnold representation theorem, using B-splines as learnable activation functions on edges. Advantages include parameter efficiency and interpretable activations. Risks are mainly that spline computation is slower than ReLU and fine grids may overfit. Suggest running ablation experiments on specific tasks.
|
|
24
31
|
|
|
25
|
-
|
|
32
|
+
**With Skill** (activated: Perturbation + Spectral lenses → `low-rank-approximation` anchor):
|
|
33
|
+
|
|
34
|
+
> **Conclusion**: The KA theorem guarantees **existence** of a representation, not **learnability**. Two overlooked mathematical quantities determine KAN's actual performance: the effective rank of the spline basis and the Fisher curvature of the parameter space.
|
|
35
|
+
>
|
|
36
|
+
> - **Existence ≠ Learnability**: The KA theorem guarantees ∑φ_q(∑ψ_{q,p}(x_p)) can represent any continuous function, but—like the universal approximation theorem for MLPs—it does not guarantee that gradient descent finds the representation
|
|
37
|
+
> - **Spline basis effective rank**: B-spline basis functions' support overlap causes effective rank to lag far behind parameter growth; if grids refine but bases remain highly correlated, actual expressive capacity stalls—monitor singular value decay curves, not just parameter counts
|
|
38
|
+
> - **Fisher curvature**: Spline's second-order differentiability makes the Hessian more regular, but the Fisher information matrix's condition number may be far larger than in ReLU networks; natural gradient has theoretical advantages but higher estimation cost
|
|
39
|
+
> - **Falsifiable experiment**: Compare KAN vs. same-parameter-budget MLP on spline basis effective rank decay; if rank growth stalls while parameters keep growing, the "interpretability" advantage is offset by redundancy
|
|
40
|
+
|
|
41
|
+
### [Crypto] Case 2: Verification — "Is this PRF watermarking reduction tight?"
|
|
26
42
|
|
|
27
|
-
|
|
43
|
+
**User**: "I use a PRF to watermark model weights. The reduction proof says 'removing watermark ≥ breaking PRF'. Is this reduction tight?"
|
|
28
44
|
|
|
29
|
-
|
|
45
|
+
**Without Skill**:
|
|
46
|
+
> If removing the watermark is equivalent to distinguishing PRF output from truly random, then the reduction is tight when the PRF security assumption holds. Suggest choosing an AES-based PRF for stronger assumptions.
|
|
47
|
+
|
|
48
|
+
**With Skill** (activated: `reduction-proof-template` + `prf-prg-owf` anchors):
|
|
49
|
+
|
|
50
|
+
> **Conclusion**: The reduction direction is correct, but tightness depends on query count Q and simulation quality—cannot claim "tight" just from "PRF assumption holds."
|
|
51
|
+
>
|
|
52
|
+
> - **Reduction loss**: Adv^scheme ≤ Q · Adv^PRF + δ, where Q is the adversary's query count. If Q grows linearly with model parameters, concrete security degrades significantly—must report Q's magnitude, not just write "polynomial loss"
|
|
53
|
+
> - **Simulation quality**: The reduction must construct simulator B to answer adversary oracle queries; if B's simulation distribution has non-negligible statistical distance from the real game, the entire reduction fails
|
|
54
|
+
> - **Assumption hierarchy**: AES-as-PRF is a widely adopted empirical assumption, not a theorem unconditionally proven from the spec—standard model, ROM, and concrete security are three levels that must not be conflated
|
|
55
|
+
> - **Multi-user degradation**: After deployment, multiple users query independently; birthday bound and hybrid steps amplify advantage loss
|
|
56
|
+
> - **Falsifiable check**: Write out B's complete simulation (parameter generation, query answering, challenge embedding, abort handling), plug in concrete parameters to estimate Q and δ
|
|
30
57
|
|
|
31
58
|
---
|
|
32
59
|
|
|
33
|
-
|
|
60
|
+
## Quick Start
|
|
34
61
|
|
|
35
|
-
|
|
62
|
+
### Installation
|
|
36
63
|
|
|
37
|
-
|
|
64
|
+
**npx** (recommended):
|
|
38
65
|
|
|
39
|
-
|
|
40
|
-
|
|
41
|
-
|
|
42
|
-
|
|
66
|
+
```bash
|
|
67
|
+
npx -y math-skill@latest install --all
|
|
68
|
+
```
|
|
69
|
+
|
|
70
|
+
**Or give the project URL to your AI and let it install itself**:
|
|
43
71
|
|
|
44
72
|
```
|
|
45
|
-
|
|
46
|
-
↓
|
|
47
|
-
Thinking Lenses: What perspective fits this problem?
|
|
48
|
-
↓
|
|
49
|
-
Activation Anchors: Which math structures to activate? Enter Knowledge Gap Protocol if insufficient
|
|
50
|
-
↓
|
|
51
|
-
Design Translation: How do these tools become model structures / losses / operators?
|
|
52
|
-
↓
|
|
53
|
-
Critic: Mathematically sound? Engineering-feasible?
|
|
73
|
+
Please help me install math-skill: https://github.com/the-thinker0/math-skill, and show me how to use it
|
|
54
74
|
```
|
|
55
75
|
|
|
56
|
-
>
|
|
76
|
+
> Also supports: `install --codex` / `--claude` (individual), `update --all` (upgrade), `doctor --all` (check for duplicates)
|
|
77
|
+
> The installer automatically excludes the nested `skills/` directory, ensuring only one entry per platform.
|
|
78
|
+
|
|
79
|
+
### Usage
|
|
80
|
+
|
|
81
|
+
**Auto-trigger**: The system auto-diagnoses user intent and routes to the appropriate layer:
|
|
82
|
+
|
|
83
|
+
| Scenario | Signal | Path |
|
|
84
|
+
|----------|--------|------|
|
|
85
|
+
| Analysis | "Is this design sound?" | Lenses → compact review |
|
|
86
|
+
| Design | "Design a new attention" | Lenses → anchors → design translation → compact review |
|
|
87
|
+
| Knowledge query | "How does tangent space relate to gradient optimization?" | Activate anchor |
|
|
88
|
+
| Verification | "Is this reduction tight enough?" | Anchors → conditions/boundaries |
|
|
89
|
+
| Pure engineering | debug, refactoring, tuning | **Not triggered** |
|
|
90
|
+
|
|
91
|
+
**Manual trigger**:
|
|
92
|
+
|
|
93
|
+
```
|
|
94
|
+
/ask <your question> # Smart diagnosis: auto-detect scenario and route
|
|
95
|
+
```
|
|
96
|
+
|
|
97
|
+
### Language
|
|
98
|
+
|
|
99
|
+
Auto-detects Chinese/English: Chinese messages get Chinese output, English messages get English output. Technical terms, code, and formulas do not determine language.
|
|
57
100
|
|
|
58
101
|
---
|
|
59
102
|
|
|
60
103
|
## Three-Layer Orthogonal Architecture
|
|
61
104
|
|
|
105
|
+
```
|
|
106
|
+
Problem → Lenses (what perspective?) → Anchors (which math structures?) → Design Translation (what module?) → Review (does it hold?)
|
|
107
|
+
```
|
|
108
|
+
|
|
62
109
|
| Layer | Role | Directory | Files |
|
|
63
110
|
|-------|------|-----------|-------|
|
|
64
111
|
| **Thinking Lenses** | Diagnose problem structure, recommend math perspectives | `lenses/*.md` | 15 |
|
|
@@ -70,7 +117,7 @@ Supporting layers:
|
|
|
70
117
|
- `references/gpu-friendly-math.en.md`: on-demand GPU checks with N/A for irrelevant dimensions
|
|
71
118
|
- `agents/math-critic.en.md`: 19-dimension deep critic loaded only for comprehensive or paper-grade review
|
|
72
119
|
|
|
73
|
-
### Domain Router
|
|
120
|
+
### Domain Router
|
|
74
121
|
|
|
75
122
|
AI research and cryptography **share** mathematical foundations (probability/information/algebra/matrix/spectrum/optimization) but each has **exclusive** specialty layers. After intent diagnosis and before lens invocation, Domain Router determines the problem's domain and decides which anchors/books/design patterns to load, avoiding cross-domain pollution and token waste.
|
|
76
123
|
|
|
@@ -129,47 +176,13 @@ AI research and cryptography **share** mathematical foundations (probability/inf
|
|
|
129
176
|
|
|
130
177
|
---
|
|
131
178
|
|
|
132
|
-
##
|
|
133
|
-
|
|
134
|
-
### Installation
|
|
135
|
-
|
|
136
|
-
```
|
|
137
|
-
Please help me install math-skill: https://github.com/the-thinker0/math-skill, and show me how to use it
|
|
138
|
-
```
|
|
139
|
-
|
|
140
|
-
Manual Codex install (install the whole repository as one self-contained skill; do not copy only the nested entry):
|
|
141
|
-
|
|
142
|
-
```bash
|
|
143
|
-
git clone https://github.com/the-thinker0/math-skill.git ~/.codex/skills/math-research-activator
|
|
144
|
-
```
|
|
145
|
-
|
|
146
|
-
Root `SKILL.md` is the canonical Codex entry. `skills/math-research-activator/SKILL.md` is only a Claude/plugin-style compatibility entry that forwards to the root. Copying the nested directory alone omits its anchors and design patterns.
|
|
147
|
-
|
|
148
|
-
### Usage
|
|
149
|
-
|
|
150
|
-
**Auto-trigger**: The system auto-diagnoses user intent and routes to the appropriate layer:
|
|
151
|
-
|
|
152
|
-
| Scenario | Signal | Path |
|
|
153
|
-
|----------|--------|------|
|
|
154
|
-
| Problem Analysis | "Is this design sound?" | Lenses → Critic |
|
|
155
|
-
| Mechanism Design | "Design a new attention" | Lenses → Anchors / Temporary Card → Design → Critic |
|
|
156
|
-
| Knowledge Query | "What is tangent space and how does it relate to gradient optimization?" | Anchors; Knowledge Gap Protocol if insufficient |
|
|
157
|
-
| Verification | "Does this formula hold?" | Anchors / Temporary Card → Critic |
|
|
158
|
-
| Pure Engineering | debug, refactoring, tuning | **Not triggered** |
|
|
159
|
-
|
|
160
|
-
**Manual trigger**:
|
|
161
|
-
|
|
162
|
-
```
|
|
163
|
-
/ask <your question> # Smart diagnosis: auto-detect scenario and route
|
|
164
|
-
```
|
|
165
|
-
|
|
166
|
-
### Language Switching
|
|
179
|
+
## Inspiration
|
|
167
180
|
|
|
168
|
-
|
|
181
|
+
The Lie group–Lie algebra machinery invented to solve differential equations ultimately became the universal language for describing symmetry and robot state estimation—the value of mathematical tools far exceeds their original intent. This is the prototype of "cross-domain activation." See [`references/inspiration.en.md`](references/inspiration.en.md).
|
|
169
182
|
|
|
170
183
|
---
|
|
171
184
|
|
|
172
|
-
##
|
|
185
|
+
## Routing Example
|
|
173
186
|
|
|
174
187
|
**User**: "Design a new KV Cache compression method that preserves long-range dependencies without just doing top-k"
|
|
175
188
|
|
|
@@ -178,25 +191,24 @@ Step 1 Diagnosis: Scenario B (Mechanism Design)
|
|
|
178
191
|
Problem type: sequence memory compression + information preservation + long-range structure
|
|
179
192
|
Core tension: compress token count vs. preserve long-range dependencies
|
|
180
193
|
|
|
181
|
-
Step 2 Lens Selection:
|
|
194
|
+
Step 2 Lens Selection (default ≤2):
|
|
182
195
|
1. Spectral (preserve dominant subspace)
|
|
183
|
-
2. Information
|
|
184
|
-
|
|
196
|
+
2. Probabilistic/Information (preserve mutual-information-sensitive states)
|
|
197
|
+
(Add Topological only when the user stresses connectivity/bridge structure; not in default budget)
|
|
185
198
|
|
|
186
199
|
Step 3 Activation Anchors:
|
|
187
200
|
→ low-rank-approximation (Matrix Analysis anchor)
|
|
188
|
-
→ leverage-score-selection (Design Pattern: compression)
|
|
189
201
|
→ information-bottleneck (Probability & Information anchor)
|
|
202
|
+
→ leverage-score-selection or low-rank-kv-cache (0–2 compression design patterns)
|
|
190
203
|
If existing anchors are insufficient, enter Knowledge Gap Protocol to generate a temporary knowledge card.
|
|
191
204
|
|
|
192
205
|
Step 4 Design Translation:
|
|
193
|
-
|
|
194
|
-
|
|
195
|
-
Candidate C: Topology-Preserving Cache (graph bridge node retention)
|
|
206
|
+
Primary: Spectral KV Compression (low-rank + leverage score)
|
|
207
|
+
Alternative (key difference only): Information-Preserving Cache (relies on future query estimation)
|
|
196
208
|
|
|
197
|
-
Step 5
|
|
198
|
-
|
|
199
|
-
|
|
209
|
+
Step 5 Compact Review (ordinary tasks do not load the full critic):
|
|
210
|
+
Primary is most GPU-friendly; alternative needs future query estimation, higher uncertainty
|
|
211
|
+
Conclusion: prioritize primary; add a lightweight gate only if query-sensitive retention is needed
|
|
200
212
|
```
|
|
201
213
|
|
|
202
214
|
---
|
|
@@ -208,14 +220,16 @@ math-skill/
|
|
|
208
220
|
├── skills/
|
|
209
221
|
│ └── math-research-activator/ # Orchestrator: intent diagnosis + routing
|
|
210
222
|
├── lenses/ # 15 thinking lenses (reasoning methodology)
|
|
211
|
-
├── knowledge-base/ # Activation anchors by math domain, not a closed encyclopedia
|
|
223
|
+
├── knowledge-base/ # Activation anchors by math domain, not a closed encyclopedia (37 cards total)
|
|
212
224
|
│ ├── matrix-analysis/ # Matrix analysis (5 cards)
|
|
213
225
|
│ ├── optimization/ # Optimization (5 cards)
|
|
214
226
|
│ ├── differential-geometry/ # Differential geometry (6 cards)
|
|
215
227
|
│ ├── lie-theory/ # Lie theory (5 cards)
|
|
216
228
|
│ ├── topology/ # Topology (3 cards)
|
|
217
229
|
│ ├── probability/ # Probability & information (5 cards)
|
|
218
|
-
│
|
|
230
|
+
│ ├── information-geometry/ # Information geometry (2 cards)
|
|
231
|
+
│ ├── algebraic-geometry/ # Algebraic geometry (2 cards)
|
|
232
|
+
│ └── cryptography/ # Cryptography (4 cards, domain-exclusive)
|
|
219
233
|
├── design-patterns/ # Design translation (by AI component)
|
|
220
234
|
│ ├── attention/ # Attention mechanisms (5 patterns)
|
|
221
235
|
│ ├── loss/ # Loss functions (5 patterns)
|
|
@@ -226,7 +240,9 @@ math-skill/
|
|
|
226
240
|
│ ├── books/ # 10 book distillations (7 AI + 3 crypto)
|
|
227
241
|
│ ├── gpu-friendly-math.en.md # GPU checklist (applicable dimensions only)
|
|
228
242
|
│ ├── agentic-workflow.en.md # Collaboration style
|
|
229
|
-
│
|
|
243
|
+
│ ├── inspiration.en.md # Inspiration
|
|
244
|
+
│ ├── musings.en.md # Musings (philosophical reflections, not auto-loaded)
|
|
245
|
+
│ └── skill-index.en.md # Index (on-demand directory, not loaded by default)
|
|
230
246
|
├── agents/math-critic.en.md # Math-engineering dual critic (19 dims, with crypto security review)
|
|
231
247
|
├── commands/ask.en.md # /ask manual entry
|
|
232
248
|
├── math_book/ # Local PDFs (not published)
|
|
@@ -249,7 +265,7 @@ math-skill/
|
|
|
249
265
|
| 6 | *An Introduction to Optimization, With Applications to ML* | Chong, Lu, Żak | John Wiley & Sons, 5th ed. | 2024 | 978-1-119-87763-9 | `optimization-ml.en.md` |
|
|
250
266
|
| 7 | *Introduction to Smooth Manifolds* | John M. Lee | Springer, GTM 218, 2nd ed. | 2013 | 978-1-4419-9981-8 | `smooth-manifolds.en.md` |
|
|
251
267
|
|
|
252
|
-
### Cryptography Direction (3 books
|
|
268
|
+
### Cryptography Direction (3 books)
|
|
253
269
|
|
|
254
270
|
| # | Title | Author(s) | Publisher / Edition | Year | ISBN | Distillation |
|
|
255
271
|
|---|-------|-----------|-------------------|------|------|-------------|
|
|
@@ -257,66 +273,58 @@ math-skill/
|
|
|
257
273
|
| 9 | *Foundations of Cryptography, Volume 1: Basic Tools* | Oded Goldreich | Cambridge University Press | 2001 | 978-0-521-79235-9 | `foundations-of-cryptography.en.md` |
|
|
258
274
|
| 10 | *Introduction to Modern Cryptography* | Jonathan Katz & Yehuda Lindell | CRC Press, 2nd ed. | 2015 | 978-1-4665-7026-1 | `introduction-to-modern-cryptography.en.md` |
|
|
259
275
|
|
|
260
|
-
> All three cryptography distillations now have Chinese and English pairs. Select `.md` or `.en.md` by the user's primary language, and load them only when anchors are insufficient or book-level depth is needed.
|
|
261
|
-
|
|
262
276
|
Distillation files ship with the npm package. For full-fidelity lookups, place PDFs in the `math_book/` folder.
|
|
263
277
|
|
|
264
278
|
---
|
|
265
279
|
|
|
266
280
|
## Changelog
|
|
267
281
|
|
|
282
|
+
### v3.3.2 — Productization & npx Installer
|
|
283
|
+
|
|
284
|
+
- **npx CLI installer**: Added `bin/math-skill.cjs` supporting `npx -y math-skill@latest install/update/doctor/uninstall`; atomic version replacement, excludes nested `skills/` to prevent double entries, `doctor` checks for duplicates
|
|
285
|
+
- **README product-first restructure**: Centered title + subtitle; demos and Quick Start on first screen (was buried at L128); Star plea moved to top; npm upgraded to npx; platforms tagged with "and more" (not framework-limited)
|
|
286
|
+
- **Case updates**: KAN analysis (AI) + PRF watermark reduction verification (Crypto) replace old FFT/orthogonality-loss cases; "without Skill" answers made realistic; case headers tagged `[AI]`/`[Crypto]`
|
|
287
|
+
- **Changelog compression**: v3.3.0–v3.1.1 compressed from 7-13 points to 5 each (~62% reduction); restored Sophus Lie inspiration section and 4-step flow diagram
|
|
288
|
+
|
|
289
|
+
### v3.3.1 — Documentation-Discipline Patch
|
|
290
|
+
|
|
291
|
+
- **README directory tree completed**: added `algebraic-geometry/`, `cryptography/`, `musings.md`, `skill-index.md`; card total aligned to 37
|
|
292
|
+
- **README workflow example corrected**: lenses default ≤2, design patterns separated from anchors, Step 5 changed to compact review — consistent with `SKILL.md` budgets
|
|
293
|
+
- **Changelog slimmed**: v1/v2 compressed to one-line summaries; removed stale "37 .en.md" / "16 thinking weapons" numbers
|
|
294
|
+
- **Entry phrasing & counting caliber unified**: three entries use consistent "when to load .en" wording; repo-wide count standardized as 33 shared + 4 crypto = 37
|
|
295
|
+
- **Metadata & hygiene**: `package.json` description/keywords trimmed; `CLAUDE.md` dir tree and Node.js note fixed; `validate.sh` gained 20 documentation-discipline checks; `original-texts` gained purpose note and v2 terminology fix
|
|
296
|
+
|
|
268
297
|
### v3.3.0 — Routing Convergence, Bilingual Completion & Technical Corrections
|
|
269
298
|
|
|
270
|
-
- **Canonical entry
|
|
271
|
-
- **
|
|
272
|
-
- **
|
|
273
|
-
- **
|
|
274
|
-
- **
|
|
275
|
-
- **Complete bilingual cryptography references**: added `applied-cryptography.en.md`, `foundations-of-cryptography.en.md`, and `introduction-to-modern-cryptography.en.md`. Commands, agents, lenses, knowledge-base, design-patterns, and references now all maintain Chinese/English pairs; only the user's primary-language side is loaded.
|
|
276
|
-
- **Relevance-driven GPU review**: candidates no longer mechanically fill all eight dimensions. Reviews first state shape, baseline, and deployment constraints, assess only decision-relevant dimensions, and mark others `N/A`. Judgments require checkable FLOPs, peak intermediate/state memory, byte or communication counts, or low-precision risks; “expressible as GEMM” is explicitly not equivalent to “faster.”
|
|
277
|
-
- **Critic and output-quality convergence**: ordinary tasks use the compact checks in the root entry, while the 19-dimension critic is reserved for paper-grade or explicitly comprehensive review. The modern-math dimension is mandatory only when a transfer claim exists; crypto prefers `knowledge-base/cryptography/` anchors before books and never uses the GPU checklist as a security gate; exploratory candidates must state assumptions, boundaries, and falsification methods.
|
|
278
|
-
- **Pre-release consistency fixes**: compatibility-entry descriptions re-synced with root `SKILL.md` (including math-query triggers); index workflow examples restored to the default ≤2-lens budget; CLAUDE/eval Gate terminology removed and the bilingual books fact corrected; sheaf-cohomology $H^1$ wording tightened so it is not confused with the sheaf axiom.
|
|
279
|
-
- **Synchronized indexes, routing examples, and language rules**: updated `skill-index`, `knowledge-base/overview`, the agentic workflow, and A/B/C/D/E eval scenarios with false-positive, intersection, knowledge-gap, and primary-language boundaries. Code, paths, formulas, and English technical terms do not vote on response language.
|
|
280
|
-
- **Evaluation-scope cleanup**: removed the six-dimension output score and token/cost regression specification created only for this audit, while retaining routing, isolation, bilingual, reference-integrity, and semantic-regression scenarios that directly test skill behavior.
|
|
281
|
-
- **Validation and package integrity**: Bash/PowerShell validation covers root frontmatter, compatibility forwarders, 37 anchors, bilingual pairing, cross-references, Domain Router isolation, the Knowledge Gap Protocol, GPU quantitative signals, and high-risk semantic regressions. The npm package explicitly includes root `SKILL*.md` and excludes PDFs, `math_book/`, tests, and the local npm cache.
|
|
299
|
+
- **Canonical entry & progressive loading**: Root `SKILL.md`/`SKILL.en.md` become self-contained normative entries; five scenarios (A/B/C/D/E) get minimal load paths (default 1-2 lenses, 1-3 anchors, 0-2 prototypes); concept queries and pure-crypto tasks no longer load full critic/books/GPU checklist by default
|
|
300
|
+
- **Domain Router rewrite**: Switched from keyword voting to "target object + requested guarantee" domain judgment; isolated terms like `hashing`/`attack`/`security` no longer trigger crypto false positives; pure-AI, pure-crypto, shared-math, and AI×crypto paths explicitly isolated
|
|
301
|
+
- **Crypto decontamination & math corrections**: Four crypto anchors reorganized around security definitions/attack games/reduction loss, no forced AI translations or GPU gates; corrected projection pseudoinverse conditions, KL directionality, low-rank gradient claims, orthogonality-loss formula; all guarantees/equivalence/optimality claims now state conditions
|
|
302
|
+
- **Bilingual & GPU review convergence**: Crypto books completed in Chinese-English pairs; GPU review switched to relevance-driven (assess only decision-relevant dimensions, mark others N/A); ordinary tasks use compact checks, 19-dim critic reserved for paper-grade review
|
|
303
|
+
- **Index, eval & validation sync**: Updated skill-index/overview/agentic-workflow and A/B/C/D/E eval scenarios; Bash/PowerShell validation covers frontmatter, 37 anchor pairs, cross-references, Domain Router isolation
|
|
282
304
|
|
|
283
305
|
### v3.2.1 — Design Philosophy Refinement & Reliability Enhancement
|
|
284
306
|
|
|
285
|
-
- **Design philosophy
|
|
286
|
-
- **
|
|
287
|
-
- **
|
|
288
|
-
- **
|
|
289
|
-
- **
|
|
290
|
-
|
|
291
|
-
|
|
292
|
-
|
|
293
|
-
|
|
294
|
-
|
|
295
|
-
**
|
|
296
|
-
|
|
297
|
-
- **
|
|
298
|
-
- `references/books/applied-cryptography.md` (Boneh & Shoup): attack games / reduction proofs / constructions / protocols
|
|
299
|
-
- `references/books/foundations-of-cryptography.md` (Goldreich): computational indistinguishability / OWF-PRG-PRF equivalence chain / simulation paradigm / meta-theorems
|
|
300
|
-
- `references/books/introduction-to-modern-cryptography.md` (Katz & Lindell): formal definitions / CPA-CCA-AE / construction paradigms / implementation pitfalls
|
|
301
|
-
- **Domain Router** (core innovation): after intent diagnosis and before lens invocation, judges problem domain (AI/Crypto/pure-math/intersection), loads domain-specific content, shared math not redundantly loaded, avoids cross-domain pollution and token waste
|
|
302
|
-
- **SKILL.md / SKILL.en.md**: new Domain Router section + routing rules + decision flow diagram; main workflow integrates domain tags and domain-specific routing (AI uses design-patterns + GPU gate; crypto uses reduction templates + assumption/pitfall checks)
|
|
303
|
-
- **math-critic upgraded to 19 dimensions**: new dim 19 "Cryptographic Security Review" (security definitions / reduction tightness / assumption dependency / composition pitfalls / anti-patterns / cross-domain transfer validity / Domain Router consistency)
|
|
304
|
-
- **skill-index / overview**: added Domain Router overview table and crypto book activation-family tags; overview gained Domain Router loading note
|
|
305
|
-
- **Token optimization**: crypto books compressed from 2084 to 404 lines (~80% reduction); Domain Router trims output by domain, avoiding full-load; output format emphasizes "after domain judgment, only expand the domain-specific subsection". Quantified estimate: pure AI problems skip the 3 crypto books entirely (saving ~400 lines/call), pure crypto problems skip all 22 AI design-patterns (saving ~2200-3300 lines/call); the crypto books' own compression saves another ~872 lines/call
|
|
306
|
-
- **File cleanup**: removed duplicate `SKILL.md/SKILL.en.md/original-texts.md/original-texts.en.md` mistakenly placed in repo root by the contributor (canonical versions live in `skills/math-research-activator/`); fixed SKILL relative paths in `agents/math-critic.{en,}.md` and `knowledge-base/overview.en.md`
|
|
307
|
-
- **AI/Cryptography isolation guarantee**: Domain Router rule 4 explicitly states "pure AI problems do not load crypto books; pure crypto problems do not load AI design patterns," preventing conceptual confusion at the loading layer
|
|
307
|
+
- **Design philosophy clarified**: Skill declared as a thinking activator, not an encyclopedia; `knowledge-base/` returns to mathematical structures (not fixed AI architectures); `design-patterns/` positioned as translation-paradigm demos, not template library
|
|
308
|
+
- **Crypto & algebraic-geometry anchors backfilled**: Added `knowledge-base/cryptography/` (4 cards) and `knowledge-base/algebraic-geometry/` (2 cards), giving Domain Router substantive crypto anchors
|
|
309
|
+
- **Testing & validation expanded**: Added scenario A/D eval, cross-domain routing, Knowledge Gap Protocol, Domain Router isolation tests; validate.sh checks card six-section structure and GPU eight-dimension coverage
|
|
310
|
+
- **Critic 19-dim layering**: Core/contextual/mandatory/meta four-tier labeling, reducing agent cognitive burden
|
|
311
|
+
- **inspiration.md split**: Technical inspiration retained; philosophical content moved to `musings.md`
|
|
312
|
+
|
|
313
|
+
### v3.2.0 — Cryptography Direction + Domain Router
|
|
314
|
+
|
|
315
|
+
- **3 crypto books added**: Boneh & Shoup, Goldreich, Katz & Lindell distillations (bilingual pairs), compressed to ~125-155 lines/book activation-index format
|
|
316
|
+
- **Domain Router routing layer**: After intent diagnosis, before lens invocation, judges problem domain (AI/crypto/pure-math/intersection), loads domain-specific content, shared math not redundantly loaded
|
|
317
|
+
- **math-critic upgraded to 19 dims**: New crypto security review dimension (security definitions/reduction tightness/assumption dependency/composition pitfalls/anti-patterns/cross-domain transfer)
|
|
318
|
+
- **Token optimization**: Crypto books compressed from 2084 to 404 lines (~80%); pure-AI problems skip crypto books (saving ~400 lines/call), pure-crypto problems skip AI design-patterns (saving ~2200 lines/call)
|
|
319
|
+
- **AI & cryptography isolation**: Domain Router explicitly states "pure-AI problems do not load crypto books; pure-crypto problems do not load AI design patterns"
|
|
308
320
|
|
|
309
321
|
### v3.1.1 — Terminology Consistency Cleanup
|
|
310
322
|
|
|
311
|
-
- **skill-index
|
|
312
|
-
- **
|
|
313
|
-
- **
|
|
314
|
-
- **README
|
|
315
|
-
- **
|
|
316
|
-
- **README activation anchors table**: column header changed from "Cards" to "Anchors"
|
|
317
|
-
- **SKILL.md / SKILL.en.md**: three-layer architecture table and intent diagnosis table updated from "Math Knowledge" to "Activation Anchors"
|
|
318
|
-
- **English README book links**: distillation files changed from `.md` to `.en.md`
|
|
319
|
-
- **validate keywords**: checks updated from "Math Knowledge" to "Activation Anchors"
|
|
323
|
+
- **"Activation anchors" caliber unified**: skill-index, README, SKILL.md "math knowledge/knowledge cards" uniformly changed to "activation anchors"
|
|
324
|
+
- **README corrections**: Workflow example step 3 from "knowledge query" to "activation anchors"; `leverage-score-selection` label corrected
|
|
325
|
+
- **SKILL.md architecture table update**: Three-layer architecture table and intent diagnosis table terminology aligned
|
|
326
|
+
- **English README book links**: Distillation files changed from `.md` to `.en.md`
|
|
327
|
+
- **validate keywords**: Changed from "math knowledge" to "activation anchors"
|
|
320
328
|
|
|
321
329
|
### v3.1.0 — Activation Anchors & Knowledge Gap Protocol
|
|
322
330
|
|
|
@@ -341,22 +349,19 @@ Distillation files ship with the npm package. For full-fidelity lookups, place P
|
|
|
341
349
|
**Architecture overhaul**: from "thinking weapon arsenal" to "math general staff" — three-layer orthogonal architecture:
|
|
342
350
|
|
|
343
351
|
- **Thinking Lenses** (15): slimmed down from v2's "thinking weapons" — reasoning methodology only, no concrete math knowledge mixed in
|
|
344
|
-
- **Knowledge Base** (31 cards): concrete math tools organized by domain, with definitions/formulas/AI design translation/GPU feasibility
|
|
352
|
+
- **Knowledge Base** (31 cards, expanded to 37 in v3.2): concrete math tools organized by domain, with definitions/formulas/AI design translation/GPU feasibility
|
|
345
353
|
- **Design Translation Layer** (new): the bridge from math to AI modules, organized by AI component (attention/loss/routing/representation/compression)
|
|
346
354
|
- **Activator rewrite**: from environment-signal matching to intent diagnosis (5 scenarios: analysis/design/query/verification/engineering)
|
|
347
355
|
- **Knowledge activation protocol**: fixed output format for knowledge cards (minimal definition → formula → applicable problems → AI translation → engineering feasibility → risks)
|
|
348
356
|
|
|
349
357
|
### v2.1.0 — Full Bilingual Support
|
|
350
|
-
-
|
|
351
|
-
|
|
352
|
-
### v2.0.1
|
|
353
|
-
- Tightened auto-trigger conditions, added exclusion gate, narrowed environment signals
|
|
358
|
+
- Full bilingual support, auto language routing, same commands for both languages, no double token cost
|
|
354
359
|
|
|
355
|
-
### v2.0.0
|
|
356
|
-
- 16 thinking weapons, modern math activation layer, GPU 8-D cross-cut
|
|
360
|
+
### v2.0.0–v2.0.1
|
|
361
|
+
- 16 thinking weapons (v2 naming, replaced by 15 lenses in v3), modern math activation layer, GPU 8-D cross-cut; tightened auto-trigger conditions and exclusion gate
|
|
357
362
|
|
|
358
363
|
### v1.0.0
|
|
359
|
-
- Initial release:
|
|
364
|
+
- Initial release: early "thinking-weapon arsenal + dual research/life paths" form (restructured into three-layer architecture in v3.0.0)
|
|
360
365
|
|
|
361
366
|
---
|
|
362
367
|
|