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/flagrare/llm-tutor/tutor-codebasenpx skills add Flagrare/llm-tutor --skill tutor-codebasegit clone --depth 1 https://github.com/Flagrare/llm-tutorWhat 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 | $0.00139 | $0.03907 |
| Opus 5 | $0.00069 | $0.01954 |
| Sonnet 5 | $0.00028 | $0.00781 |
| Haiku 4.5 | $0.00014 | $0.00391 |
Grade B, and why
tutor-codebase scanned grade B with 1 finding 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.
Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
For overview aspect: *"Have you worked in this codebase before? Read the README, poked at any files, run any commands?"* How it starts
The opening of the file, as written. The whole thing — 300 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tutor Codebase
The distinctive skill of llm-tutor. While /tutor-start teaches abstract topics from world knowledge, /tutor-codebase teaches this specific codebase by reading actual files and building a learning path grounded in real code. Per the research catalog, no commercial LLM tutor handles "teach me this codebase" — that's our wedge.
The mechanics parallel /tutor-start (calibration, cycle charge, persona handoff) but with three real differences:
- Pre-path exploration. The tutor reads files (
Read/Grep/Glob) before generating the path. The path's concept ordering reflects the real architecture, not generic textbook ordering. - File-anchored concepts. Each concept has the format
"<abstract framing> (<file>:<symbol>)". The tutor knows what to reference during the dialogue. - Calibration probes familiarity with THIS codebase. Not "do you know auth in general" — "have you read
src/auth/middleware.tsbefore?"
Zero deps: this skill uses Read/Grep/Glob (built-in Claude Code tools) for all codebase exploration. It does NOT call out to other plugins' codebase-exploration skills.
Step 0 — Pre-checks
STATE="$CLAUDE_PLUGIN_ROOT/scripts/state.sh"
TIER="$CLAUDE_PLUGIN_ROOT/scripts/tier.sh"
0a. Initialize state + refill cycles
bash "$STATE" init >/dev/null
bash "$STATE" refill-cycles >/dev/null 2>&1 || true
CYCLES=$(bash "$STATE" get .user.cycles)
If $CYCLES < 1, refuse:
"Out of cycles — they refill once per day. Run
/tutor-statusto see when next refill is due."
Exit.
0b. Resolve PATH argument
The user invokes /tutor-codebase <path> or /tutor-codebase <path> "aspect description".
- The
<path>is required. Without it, refuse: "Need a path. Try/tutor-codebase .for the current directory, or/tutor-codebase ./authfor a subdirectory." - Validate that the path exists and is a directory. If not: "
<path>isn't a directory I can read." Exit.
Resolve to absolute path:
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 · 300 lines · 139 tokens per session scan B d0e847b0c978
tutor-codebase is a skill published in the GitHub repository Flagrare/llm-tutor (5 stars, last pushed 3mo ago), licensed MIT. It adds 139 tokens to every session and 3,907 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ideogram-ultra
Build Ideogram 4 (Ideogram Ultra) txt2img and img2img workflows with the local open-weights model, dual conditional/unconditional models with DualModelGuider, Qwen3-VL text encoder, and structured JSON ("compositional deconstruction") prompts for strong text rendering and layout control.
model-compatibility
Model family compatibility matrix covering loaders, resolutions, samplers, CFG, VAE, ControlNet, and LoRA compatibility for SD 1.5, SDXL, Flux, SD3, and video models.
flux-txt2img
Build Flux txt2img workflows with Flux.1 Dev (SRPO), Flux 2 Klein 9B, Turbo LoRAs, FluxGuidance, and DualCLIPLoader patterns.
panel-node-pack-sync
Keep the ComfyUI sidebar panel node-pack (comfyui-agent-panel) in step with the orchestrator after comfyui-mcp updates. Use this whenever the orchestrator was just updated (installcomfyui (action:"selfupdate"), npm i -g comfyui-mcp, a new version in the ENVIRONMENT line), when a panel/bridge command fails in a way…
civitai
Discover Civitai models with the BUILT-IN downloadmodel action:"searchcivitai" and install/generate them locally. Find a checkpoint/LoRA/embedding on Civitai, download it into ComfyUI, and use its trigger words. Optionally pair the official Civitai MCP for community features (images browsing, posting, collections).
20-client-intake-brief-global
Use when an agency or freelancer takes on a NEW client and needs everything up front — an 11-section intake covering business overview, audience, current marketing, goals, budget, timeline, brand assets, approval process, and success criteria, written for the client to fill in and return. Trigger on 'client intake'…