Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/nvk/llm-wiki/ideagit clone --depth 1 https://github.com/nvk/llm-wikiWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00018 | $0.01007 |
| Opus 5 | $0.00009 | $0.00504 |
| Sonnet 5 | $0.00004 | $0.00201 |
| Haiku 4.5 | $0.00002 | $0.00101 |
Grade A, and why
idea scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Your task
Manage Ideas between compiled Concepts and committed Projects. Read
skills/wiki-manager/references/ideas.md; load inventory, research, or Project
references only for the operation that needs them.
Resolve HUB from $HOME/.config/llm-wiki/config.json (hub_path first), then
--local, --wiki, CWD .wiki/, or HUB. Read indexes first. Writes require
one active topic and never go at hub root.
Natural language is primary. Infer new, list, show, develop, shape,
promote, or archive; resolve phrases such as "the podcast thing" from the
conversation, titles, aliases, and Idea indexes. Derive safe slugs.
Capture: new [<slug>] "<seed>"
Check Idea titles/aliases, relevant Concepts/inventory, and Project slugs for duplicates. Link related Ideas; never merge automatically.
- Lazily create
inventory/ideas/and its_index.md. - Write
inventory/ideas/<slug>.mdwith required inventory fields,kind: idea,status: proposed, defaultpriority: p2,next_action, and related wiki/source links when known. - Preserve the exact seed under
## Original Seed. Add concise sections for opportunity, target, evidence, assumptions/counterevidence, alternatives, current shape, scope/no-gos, risks, questions, success criteria, intended deliverable, and decision notes. - Rebuild indexes and log
idea | captured <slug>.
Do not duplicate the seed as a raw note or create a Project.
Catalog: list and show
Read the Idea index first; hub-wide lists read only active topics' Idea indexes. Filter by status, maturity, priority, topic, tags/aliases, owner, next action, or Project. Default to a short list.
Derive maturity: captured when the record exists; researching from linked raw
research; shaped from approved_at plus an Approved Decision Snapshot; project
from a linked non-empty WHY.md; delivered from a linked outcome/deliverable.
Never add a synchronized stage field.
Develop: develop <slug>
Read the Idea, linked Concepts, and indexes; research only gaps. Evidence goes
in raw/, reusable synthesis in wiki/, and the Idea distinguishes evidence,
assumptions, contradictions, and unknowns. Update status/date/next action,
indexes, and log. Research never implies approval or Project creation.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 89 lines · 18 tokens per session scan A b83d93f1e6ff
idea is a command published in the GitHub repository nvk/llm-wiki (1,175 stars, last pushed 5d ago), licensed MIT. It adds 18 tokens to every session and 1,007 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
feedback
Collect feedback and create GitHub issue (bug report, feature request).
todo
Backlog queue for Kanban Mode — add, list, or pick the next card.
story-implement
Implement a story through the Implement Phase. Stories must be fully designed with chunks before implementation.
planning
Feature roadmap and planning. Manage initiatives, concepts, open questions, and story creation from planning.
story-new
Create a new story. It lands in the backlog until assigned to a cycle.
cycle-complete
Complete a cycle. Triggers reflection if pending learnings, then archives.