A local learning lab for organising study topics and completing private study sessions. It works like JupyterLab, a browser-based workspace for interactive learning, while keeping data on your machine.

Use it to read syllabi, create study units, choose a topic, track completions, save notes and links, and review the available study pool.

It turns syllabi into focused study units and keeps the study library, progress, and runtime data local. An AI assistant can operate the same store through files, the command line, or MCP.

rolldeep — Skill by dev-the-dev-while-deving

Install

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.

agentmods

npx agentmods add skills/dev-the-dev-while-deving/rolldeep-lab/rolldeep-lab

Any agent

npx skills add dev-the-dev-while-deving/rolldeep-lab --skill rolldeep-lab

Clone the repo

git clone --depth 1 https://github.com/dev-the-dev-while-deving/rolldeep-lab

Made for: Claude Code, Codex.

Per session 60 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 536 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What 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	$0.00060	$0.00536
Opus 5	$0.00030	$0.00268
Sonnet 5	$0.00012	$0.00107
Haiku 4.5	$0.00006	$0.00054

Measured yesterday against content hash bd352b772e63, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

rolldeep 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 yesterday.

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.

SKILL.md · 69 lines

What it actually says

RollDeep harness

You are the backend. The browser is the roll surface. Data stays on the machine (~/.rolldeep/). The pool lives in repo files.

Activate (human)

From the repo, after npm install:

rolldeep

Same idea as jupyter lab. Aliases: rolldeep lab, rolldeep notebook.

Activate (any model)

Prefer files, then CLI. MCP if configured.

Read syllabi in content/syllabi/.
Write units to content/units/<name>.json.
Sync.

{
  "title": "Course name",
  "units": [
    {
      "title": "Syllabus topic name",
      "question": "One sharp curiosity question. Never What is X?",
      "minutes": 40,
      "stars": 3,
      "sourceExcerpt": "module / section"
    }
  ]
}

rolldeep sync
rolldeep syllabi
rolldeep use --title "Course name"
rolldeep status
rolldeep roll
rolldeep choose --id <id>
rolldeep complete --id <id> --url https://... --notes "..."

MCP stdio: rolldeep mcp

Tools: rolldeep_status, rolldeep_roll, rolldeep_choose, rolldeep_complete, rolldeep_library, rolldeep_pool, rolldeep_syllabi, rolldeep_use, rolldeep_add_units, rolldeep_sync, rolldeep_delete_available.

Rules

Topic name and curiosity question are separate fields.
Max 5 rolls per session. Only current + 3 previous stay visible.
TAKE (choose) before complete. Proof URL required.
Completed titles never return on that syllabus. Other syllabi keep their own progress.
One content/units/*.json file is one syllabus. Switch with the lab dropdown or rolldeep use.
Do not invent study templates or filming scripts.
Do not call an extra model to mint units. Author them into the JSON files.

Changes

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.

yesterday First seen · 69 lines · 60 tokens per session scan A bd352b772e63

Subscribe to this mod's changes

rolldeep is a skill published in the GitHub repository dev-the-dev-while-deving/rolldeep-lab (0 stars, last pushed 11d ago), licensed MIT. It adds 60 tokens to every session and 536 once invoked, about $0.0003 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 2026-08-31.

Learning and teaching jupyter learning local-first mcp

rolldeep

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

RollDeep harness

Activate (human)

Activate (any model)

Rules

What this file has done since we first saw it

Other skills, from other repositories

learn-patchquest

trulens-notebook-execution

reading-metaskill

coach

clinical-q-and-a

guideline-section-expansion