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 skills add bearded-illirian/trailmark --skill note-firstgit clone --depth 1 https://github.com/bearded-illirian/trailmarkWrote 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/bearded-illirian/trailmark/note-first)<a href="https://agentmods.dev/skills/bearded-illirian/trailmark/note-first"><img src="https://agentmods.dev/badge/skills/bearded-illirian/trailmark/note-first/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/bearded-illirian/trailmark/note-first"><img src="https://agentmods.dev/badge/skills/bearded-illirian/trailmark/note-first.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00066 | $0.00623 |
| Opus 5.5 | $0.00026 | $0.00249 |
| Sonnet 5.5 | $0.00013 | $0.00125 |
| Haiku 4.5 | $0.00007 | $0.00062 |
Grade A, and why
note-first 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 1mo 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 — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Note-First Protocol
Saves the last Claude message as a note attached to the current task.
Input
The user's request to save the agent's most recent chat message; the current task context.
Output
note-{NN}.md with two-way links (task.md ↔ note file), auto-numbered.
Hands off to
— (terminal).
Step 0 — Task context
Ensure {log_dir} and {slug} are known. Otherwise resolve from the latest task log folder.
Cannot skip — without {log_dir} the note has nowhere to go.
Step 1 — Capture the last message
Take the assistant message that comes immediately before the /note-first command.
Rules:
- Copy verbatim — no abbreviations, no rephrasing
- If no previous assistant message exists → tell the user, finish
Step 2 — Determine note number
ls {log_dir}/note-*.md 2>/dev/null | sort | tail -1
- No files →
note-01.md note-02.mdexists → writenote-03.md
Step 3 — Write file + update task.md
Write the note
> 🔗 Task: tasks/log/{slug}/task.md
# Note NN
**Date:** {date}
{verbatim last assistant message}
Update task.md
Add a header line after the last > 📋 or > 📝 line:
> 📝 Note N: tasks/log/{slug}/note-NN.md
If none exist — add as the first line of the file.
Step 4 — Confirm
Report in one line:
📝 Saved: {log_dir}/note-NN.md
Nothing more — the note is already visible in the chat.
Anti-patterns
❌ Rephrasing before saving
Copy the message verbatim. The user is saving what was actually said.
❌ Writing the file but not updating task.md
Both actions in Step 3 are mandatory. Otherwise the note is orphaned.
❌ Saving the user's own message
The /note-first command is not the content. Save the previous assistant reply.
❌ Imitating the skill inline
Always via $note-first. Never copy the protocol from memory — outdated versions drift silently.
Related skills
- decision-first — saves an architectural decision (structured, not free-form)
- plan-first — plans execution steps (before code, not after)
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.
- 1mo ago Changed 32fb3cbb589f
- 1mo ago First seen · 117 lines · 66 tokens per session scan A 14cccf6e340f
note-first is a skill published in the GitHub repository bearded-illirian/trailmark (19 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 623 once invoked, about $0.0003 per session on Opus 5.5. 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
agent-self-scheduling
Schedule AI agent runs with cron, loops, or external clocks while avoiding unsafe tight autonomous timers.
agentmail
Use your assigned AgentMail inbox to read email tasks, explicitly send or reply, and check delivery. Provided automatically by your inbox assignment.
google-calendar
Read schedules, check availability, and create or update Google Calendar events when the user asks to work with their calendar.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
mission-slice-sop
Use when starting, building, handing off, restoring, or closing a mission or slice under the default lightweight SDLC.
material-organizer
Organize only the materials the user has already provided or explicitly pointed to. The job is to make messy inputs easier to scan, reuse, archive, or continue from later while preserving source traceability.