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 skills/janyork/llm-wiki-cli/using-practicenpx skills add JanYork/llm-wiki-cli --skill using-practicegit clone --depth 1 https://github.com/JanYork/llm-wiki-cliWhat 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 | $0.00044 | $0.00650 |
| Opus 5 | $0.00022 | $0.00325 |
| Sonnet 5 | $0.00009 | $0.00130 |
| Haiku 4.5 | $0.00004 | $0.00065 |
Grade A, and why
using-practice 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 3d 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Using Practice
Practice preserves assessed work across context loss. Flashcards are Practice items. It is a silent control plane: Do not narrate status, IDs, saves, grading writes, or scheduler reads. Show only questions, answers, feedback, and meaningful progress.
Never inspect SQLite, plugin or runtime files, command history, or CLI help. Never reread Skill files or text-search for state. Use only the public commands here.
Enter or recover
Do not load Practice for a Tutor A/B reply, ordinary comprehension check, or lightweight diagnostic. Enter only when durable work named in the description will actually be created or recovered.
- Inspect
LWC_READINESS.practice. If disabled, explain the durable local-data boundary and ask once beforelwc --scope global config set --practice enabled; explicit enablement is consent. Enabling does not download; first status may lazily install the pinned, hash-verified runtime. - Run
lwc practice statusonce on entry, lost identity, compaction, or a pending/revision/owner error. Otherwise cache exact attempt/session/owner/latest revision; do not repeat status. - Recover an in-progress attempt and saved responses before creating another. Use a
new stable
request_idper operation; retry that operation with the same request ID, owner, andif_revision. - Cross-machine recovery requires the latest successful Sync receipt and exact-revision takeover; the old owner must stop writing.
Durable flow
- Resolve exact typed subject, goal, book, bank, set, paper, and item IDs. Precedence is explicit plan target, goal bank, then subject default. Invalid higher-precedence targets fail; never fuzzy-match.
- Generated items stay draft until prompt, answer, rubric, and exact source ref are re-read and verified. Formal papers use verified revisions and report shortages without weakening the blueprint.
- Start or resume the exact attempt. Save every response immediately with owner,
request ID, and
if_revision; submission freezes already durable responses. Abandonment preserves history. v1 accepts only choice, text, numeric, and flashcard self-rating; reject every other form. - Grade objective items deterministically. Grade subjective items against the frozen rubric and store concise rationale, confidence, method, and review state. Learner overrides append history.
- Let the pinned scheduler compute review state and due dates. Respect time budget and retention target, expose deferred debt, and never edit due dates or silently control review.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 3d ago First seen · 57 lines · 44 tokens per session scan A ba1997757ebb
using-practice is a skill published in the GitHub repository JanYork/llm-wiki-cli (48 stars, last pushed 7d ago), licensed Apache-2.0. It adds 44 tokens to every session and 650 once invoked, about $0.0002 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 skills, from other repositories
openspec-onboard
Guided onboarding for OpenSpec - walk through a complete workflow cycle with narration and real codebase work.
book-to-skill
Converts books and documents (PDF, EPUB, DOCX, HTML, Markdown, plain text, RTF, MOBI/AZW with Calibre) into structured agent skills, extracting frameworks, mental models, principles, techniques, and anti-patterns. Use when the user wants to study a document through GitHub Copilot CLI, Amp, Claude Code, or Hermes…
cortex-remember
Store important decisions, patterns, errors, lessons, and context into Cortex persistent memory. Use when the user says 'remember this', 'save this', 'store this for later', 'note this down', 'don't forget', 'this is important', 'bookmark this', or when a significant decision, bug fix, architecture choice, or lesson…
learn-content
Ingest URLs, articles, or text from any source into structured knowledge cards. Stores source URL + metadata + briefing (NOT full text). On-demand retrieval from source. TRIGGER: "learn this", "save this article", "read and remember", "ingest this". NOT FOR: deep-research, summarize, persist use cases.
teach-me
Turn a 'teach me X' request into a single interactive HTML lesson, rendered Arcade-first, by emitting a lesson data model and assembling it with rp1 tooling.
socratic-duel-run
Run Socratic Duel through two participant subagents while the launcher only coordinates and reports.