Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/KaydenClark/LLM_Workbenchnpx agentmods add skills/kaydenclark/llm_workbench/notepadWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kaydenclark/llm_workbench/notepad)<a href="https://agentmods.dev/skills/kaydenclark/llm_workbench/notepad"><img src="https://agentmods.dev/badge/skills/kaydenclark/llm_workbench/notepad/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kaydenclark/llm_workbench/notepad"><img src="https://agentmods.dev/badge/skills/kaydenclark/llm_workbench/notepad.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 5.1 | $0.00056 | $0.02221 |
| Opus 5 | $0.00028 | $0.01111 |
| Sonnet 5 | $0.00011 | $0.00444 |
| Haiku 4.5 | $0.00006 | $0.00222 |
Grade A, and why
notepad 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 today.
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 — 201 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A notepad is the working context of one objective: what the owner directed, what you found and where, what you corrected, what is still unresolved, and what you would do next. It exists so that a fresh reader - you after a compaction, another agent, another provider - can continue without the owner reconstructing the conversation.
It is not evidence and not authority. A note never proves a claim, never authorizes an action, and never overrides the Contract. On resume, obey the current controls and verify live state; the note tells you what to look at, not what is true.
workbench/tools/notepads.mjs owns the structure - schema, safe revision-checked
writes, discovery, bounded retrieval, dependency-preserving cleanup. This skill
owns the judgment: what is worth saving, and when it is safe to remove.
1. Create or resume before the context is only in the conversation
Meaningful objective work needs a note; trivial conversation does not. The test is survival value: would losing this impair continuation or a focused handoff?
Resume by the strongest available signal - an explicit note the owner named, then the objective key, then the most recently created local note - and check relevance before using it:
node workbench/tools/notepads.mjs list --objective OBJECTIVE_KEY
node workbench/tools/notepads.mjs read --note NOTE --view current
--view current returns the compact resumption view and the revision to write
against, without putting the entry history into your response. Start there.
Create only when no existing note owns the objective:
node workbench/tools/notepads.mjs create --note NAME --objective OBJECTIVE_KEY \
--title "TITLE" --focus "WHAT THIS NOTE IS FOR"
Live notes stay local and untracked, in a live collection workbench/manifest.json
declares. Bare names use notepads/work/ on the new layout, or the legacy
grilling collection when that room has not migrated. Explicit relative paths
select other local JSON-notepad type folders. Handoffs are not JSON records:
write them as Markdown in sessions/handoffs/. Schema and examples under
notepads/templates/ are tracked and cannot be live
notes. Never
commit one, and never cite one as durable evidence.
For a new visible identity, use notepads.mjs allocate --prefix N with the
same objective/title fields. The returned ID is the
visible label and filename; no second identity is added. Read an allocated or
legacy visible ID through --id ID, even when its existing filename differs.
Use --note for the original filename/path lookup. Use one writer, preserve existing paths, and reconcile unreadable or ambiguous
inventory before allocating. The Runbook owns alphabet, width and collision
rules; neither an ID nor allocation grants authority.
Before dependent work, name the active note and returned revision in local execution state, verify relevant Actuality, and correct any stale current view. An unavailable runtime blocks the dependent capture operation; report it and continue only independent work whose context can safely be preserved.
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.
- today Changed · +23 lines afb9109d4a7b
- yesterday First seen · 178 lines · 56 tokens per session scan A f5b52e6176eb
notepad is a skill published in the GitHub repository KaydenClark/LLM_Workbench (2 stars, last pushed today), licensed MIT. It adds 56 tokens to every session and 2,221 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 2026-09-09.
Other skills, from other repositories
media-ingest
Ingest video, audio, PDF, book, screenshot, and GitHub repo content into the brain. Multi-format handling with entity extraction and backlink propagation. Covers video-ingest, youtube-ingest, and book-ingest subtypes.
mem0-oss-to-platform
Plan and then execute a migration of a project from the mem0 open-source / self-hosted SDK (the local Memory class) to the mem0 Platform / hosted / managed SDK (the MemoryClient class). Use this whenever a developer wants to move, switch, or migrate their mem0 usage off OSS/self-hosted to the hosted API — e.g.…
Cortex
Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…
memory
Use when the user asks to remember, recall, forget, update, search, or inspect durable OpenSquilla memory, including profile facts in USER.md and long-term notes in MEMORY.md or memory//.md.
ha-data-stores
Map of Hope Agent's local data stores and safe read-only query workflow. Use when the user asks where Hope Agent stores data, wants to inspect sessions/messages/memory/logs/background jobs/knowledge indexes/settings, asks the model to query local app data, or debugging requires checking persisted state. Trigger…
shellm
Reference for the shellm system — recursive LLM shell, identity management, memory, skills, trajectory, and all CLI tools. Use when working on shellm itself, debugging agent behavior, or understanding how the pieces fit together.