llm-wiki

Name: llm-wiki
Author: adityahimaone

72stars5forks

How to install llm-wiki?

adityahimaone/hermes-agent-rtk-caveman/llm-wiki

$npx -y skills add adityahimaone/hermes-agent-rtk-caveman --skill llm-wiki

Installs into the current project.

Summary

Karpathy's LLM Wiki — build and maintain a persistent, interlinked markdown knowledge base. Ingest sources, query compiled knowledge, and lint for consistency.

Files · 1

View on GitHub

SKILL.md

1	# Karpathy's LLM Wiki
2
3	Build and maintain a persistent, compounding knowledge base as interlinked markdown files.
4	Based on [Andrej Karpathy's LLM Wiki pattern](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f).
5
6	Unlike traditional RAG (which rediscovers knowledge from scratch per query), the wiki
7	compiles knowledge once and keeps it current. Cross-references are already there.
8	Contradictions have already been flagged. Synthesis reflects everything ingested.
9
10	Division of labor: The human curates sources and directs analysis. The agent
11	summarizes, cross-references, files, and maintains consistency.
12
13	## When This Skill Activates
14
15	Use this skill when the user:
16	- Asks to create, build, or start a wiki or knowledge base
17	- Asks to ingest, add, or process a source into their wiki
18	- Asks a question and an existing wiki is present at the configured path
19	- Asks to lint, audit, or health-check their wiki
20	- References their wiki, knowledge base, or "notes" in a research context
21
22	## Wiki Location
23
24	Configured via `skills.config.wiki.path` in `~/.hermes/config.yaml` (prompted
25	during `hermes config migrate` or `hermes setup`):
26
27	```yaml
28	skills:
29	config:
30	wiki:
31	path: ~/wiki
32	```
33
34	Falls back to `~/wiki` default. The resolved path is injected when this
35	skill loads — check the `[Skill config: ...]` block above for the active value.
36
37	The wiki is just a directory of markdown files — open it in Obsidian, VS Code, or
38	any editor. No database, no special tooling required.
39
40	## Architecture: Three Layers
41
42	```
43	wiki/
44	├── SCHEMA.md # Conventions, structure rules, domain config
45	├── index.md # Sectioned content catalog with one-line summaries
46	├── log.md # Chronological action log (append-only, rotated yearly)
47	├── raw/ # Layer 1: Immutable source material
48	│ ├── articles/ # Web articles, clippings
49	│ ├── papers/ # PDFs, arxiv papers
50	│ ├── transcripts/ # Meeting notes, interviews
51	│ └── assets/ # Images, diagrams referenced by sources
52	├── entities/ # Layer 2: Entity pages (people, orgs, products, models)
53	├── concepts/ # Layer 2: Concept/topic pages
54	├── comparisons/ # Layer 2: Side-by-side analyses
55	└── queries/ # Layer 2: Filed query results worth keeping
56	```
57
58	Layer 1 — Raw Sources: Immutable. The agent reads but never modifies these.
59	Layer 2 — The Wiki: Agent-owned markdown files. Created, updated, and
60	cross-referenced by the agent.
61	Layer 3 — The Schema: `SCHEMA.md` defines structure, conventions, and tag taxonomy.
62
63	## Resuming an Existing Wiki (CRITICAL — do this every session)
64
65	When the user has an existing wiki, always orient yourself before doing anything:
66
67	① Read `SCHEMA.md` — understand the domain, conventions, and tag taxonomy.
68	② Read `index.md` — learn what pages exist and their summaries.
69	③ Scan recent `log.md` — read the last 20-30 entries to understand recent activity.
70
71	```bash
72	WIKI="${wiki_path:-$HOME/wiki}"
73	# Orientation reads at session start
74	read_file "$WIKI/SCHEMA.md"
75	read_file "$WIKI/index.md"
76	read_file "$WIKI/log.md" offset=<last 30 lines>
77	```
78
79	Only after orientation should you ingest, query, or lint. This prevents:
80	- Creating duplicate pages for entities that already exist
81	- Missing cross-references to existing content
82	- Contradicting the schema's conventions
83	- Repeating work already logged
84
85	For large wikis (100+ pages), also run a quick `search_files` for the topic
86	at hand before creating anything new.
87
88	## Initializing a New Wiki
89
90	When the user asks to create or start a wiki:
91
92	1. Determine the wiki path (from config, env var, or ask the user; default `~/wiki`)
93	2. Create the directory structure above
94	3. Ask the user what domain the wiki covers — be specific
95	4. Write `SCHEMA.md` customized to the domain (see template below)
96	5. Write initial `index.md` with sectioned header
97	6. Write initial `log.md` with creation entry
98	7. Confirm the wiki is ready and suggest first sources to ingest
99
100	### SCHEMA.md Template
101
102	Adapt to the user's domain. The schema constrains agent behavior and ensures consistency:
103
104	```markdown
105	# Wiki Schema
106
107	## Domain
108	[What this wiki covers — e.g., "AI/ML research", "personal health", "startup intelligence"]
109
110	## Conventions
111	- File names: lowercase, hyphens, no spaces (e.g., `transformer-architecture.md`)
112	- Every wiki page starts with YAML frontmatter (see be

People who liked this also viewed

6 picks

Type

azure-aiUse for Azure AI: Search, Speech, OpenAI, Document Intelligence. Helps with search, vector/hybrid search, speech-to-text, text-to-speech, transcription, OCR. WHEN: AI Search, query search, vector searSkillsJul 2026472k1.3k
just-scrapeSearch, scrape, crawl, extract structured data, and monitor web pages via the ScrapeGraph AI CLI. Use when the user asks to search the web, scrape a webpage, grab content from a URL, extract JSON fromSkillsJul 2026245k37
firecrawlCLI and Agent Skill for Firecrawl - Add scrape, search, and browsing capabilities to your AI agentsSkillsJul 202694k531
firecrawl-searchWeb search with full page content extraction. Use this skill whenever the user asks to search the web, find articles, research a topic, look something up, find recent news, discover sources, or says "SkillsJul 202676k531
firecrawl-agentAI-powered autonomous data extraction that navigates complex sites and returns structured JSON. Use this skill when the user wants structured data from websites, needs to extract pricing tiers, producSkillsJul 202674k531
firecrawl-mapDiscover and list all URLs on a website, with optional search filtering. Use this skill when the user wants to find a specific page on a large site, list all URLs, see the site structure, find where sSkillsJul 202674k531

llm-wiki

72stars5forks

View on GitHub

How to install llm-wiki?

adityahimaone/hermes-agent-rtk-caveman/llm-wiki

$npx -y skills add adityahimaone/hermes-agent-rtk-caveman --skill llm-wiki

Installs into the current project.

Summary

Karpathy's LLM Wiki — build and maintain a persistent, interlinked markdown knowledge base. Ingest sources, query compiled knowledge, and lint for consistency.

Files · 1

View on GitHub

SKILL.md

1	# Karpathy's LLM Wiki
2
3	Build and maintain a persistent, compounding knowledge base as interlinked markdown files.
4	Based on [Andrej Karpathy's LLM Wiki pattern](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f).
5
6	Unlike traditional RAG (which rediscovers knowledge from scratch per query), the wiki
7	compiles knowledge once and keeps it current. Cross-references are already there.
8	Contradictions have already been flagged. Synthesis reflects everything ingested.
9
10	Division of labor: The human curates sources and directs analysis. The agent
11	summarizes, cross-references, files, and maintains consistency.
12
13	## When This Skill Activates
14
15	Use this skill when the user:
16	- Asks to create, build, or start a wiki or knowledge base
17	- Asks to ingest, add, or process a source into their wiki
18	- Asks a question and an existing wiki is present at the configured path
19	- Asks to lint, audit, or health-check their wiki
20	- References their wiki, knowledge base, or "notes" in a research context
21
22	## Wiki Location
23
24	Configured via `skills.config.wiki.path` in `~/.hermes/config.yaml` (prompted
25	during `hermes config migrate` or `hermes setup`):
26
27	```yaml
28	skills:
29	config:
30	wiki:
31	path: ~/wiki
32	```
33
34	Falls back to `~/wiki` default. The resolved path is injected when this
35	skill loads — check the `[Skill config: ...]` block above for the active value.
36
37	The wiki is just a directory of markdown files — open it in Obsidian, VS Code, or
38	any editor. No database, no special tooling required.
39
40	## Architecture: Three Layers
41
42	```
43	wiki/
44	├── SCHEMA.md # Conventions, structure rules, domain config
45	├── index.md # Sectioned content catalog with one-line summaries
46	├── log.md # Chronological action log (append-only, rotated yearly)
47	├── raw/ # Layer 1: Immutable source material
48	│ ├── articles/ # Web articles, clippings
49	│ ├── papers/ # PDFs, arxiv papers
50	│ ├── transcripts/ # Meeting notes, interviews
51	│ └── assets/ # Images, diagrams referenced by sources
52	├── entities/ # Layer 2: Entity pages (people, orgs, products, models)
53	├── concepts/ # Layer 2: Concept/topic pages
54	├── comparisons/ # Layer 2: Side-by-side analyses
55	└── queries/ # Layer 2: Filed query results worth keeping
56	```
57
58	Layer 1 — Raw Sources: Immutable. The agent reads but never modifies these.
59	Layer 2 — The Wiki: Agent-owned markdown files. Created, updated, and
60	cross-referenced by the agent.
61	Layer 3 — The Schema: `SCHEMA.md` defines structure, conventions, and tag taxonomy.
62
63	## Resuming an Existing Wiki (CRITICAL — do this every session)
64
65	When the user has an existing wiki, always orient yourself before doing anything:
66
67	① Read `SCHEMA.md` — understand the domain, conventions, and tag taxonomy.
68	② Read `index.md` — learn what pages exist and their summaries.
69	③ Scan recent `log.md` — read the last 20-30 entries to understand recent activity.
70
71	```bash
72	WIKI="${wiki_path:-$HOME/wiki}"
73	# Orientation reads at session start
74	read_file "$WIKI/SCHEMA.md"
75	read_file "$WIKI/index.md"
76	read_file "$WIKI/log.md" offset=<last 30 lines>
77	```
78
79	Only after orientation should you ingest, query, or lint. This prevents:
80	- Creating duplicate pages for entities that already exist
81	- Missing cross-references to existing content
82	- Contradicting the schema's conventions
83	- Repeating work already logged
84
85	For large wikis (100+ pages), also run a quick `search_files` for the topic
86	at hand before creating anything new.
87
88	## Initializing a New Wiki
89
90	When the user asks to create or start a wiki:
91
92	1. Determine the wiki path (from config, env var, or ask the user; default `~/wiki`)
93	2. Create the directory structure above
94	3. Ask the user what domain the wiki covers — be specific
95	4. Write `SCHEMA.md` customized to the domain (see template below)
96	5. Write initial `index.md` with sectioned header
97	6. Write initial `log.md` with creation entry
98	7. Confirm the wiki is ready and suggest first sources to ingest
99
100	### SCHEMA.md Template
101
102	Adapt to the user's domain. The schema constrains agent behavior and ensures consistency:
103
104	```markdown
105	# Wiki Schema
106
107	## Domain
108	[What this wiki covers — e.g., "AI/ML research", "personal health", "startup intelligence"]
109
110	## Conventions
111	- File names: lowercase, hyphens, no spaces (e.g., `transformer-architecture.md`)
112	- Every wiki page starts with YAML frontmatter (see be

People who liked this also viewed

6 picks

Type