Chalk Labs
Need help deploying vigil?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/piromagnus-vigil
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Nov 2, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/vigil
What's inside
- Skill name
vigil- Description
- The Vigil, the welcome conversation of a new household. The Seneschal asks the user one question at a time, in six chapters, and writes each answer in the workspace at once. Use when BOOTSTRAP.md exists, when the user says "start the vigil" or "continue the vigil", or when `korrigans context` is below 100% and the user has time.
- License
- AGPL-3.0
- Instructions
- 50 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- files: referenced but missing: templates/person.md, templates/project.md
Files (1)
- SKILL.md
SKILL.md (first lines)
# The Vigil
Before being dubbed, a knight keeps a vigil of arms. Here, the Seneschal keeps one with the user: a relaxed
conversation that gives the household everything it needs to work, chapter by chapter. It should feel like
the opening of a game, not a form.
## How to run it
1. Run `korrigans context` and start at the first chapter that is not complete. Show the bars at the start
and after each chapter: the user sees the household fill up.
2. Ask **one question at a time**, in the user's language, with 2 to 4 suggested answers the user can pick
(number them), plus "something else". Use what install detected (`memory/global/detected.md`) to
suggest: "I found `~/code/field-tools`: is it one of your projects?"
3. **Write each answer right away** in the file listed below, then remove the matching `vigil:` mark from
that file. Never keep an answer only in the conversation.
4. At the end of each chapter: one-line recap of what you wrote, `git add -A && git commit -m "vigil: <chapter>"`
in the workspace (if it is a git repository), then `korrigans context`.
5. **Real work comes first.** If the user brings a real task mid-way, handle it (or delegate it), then offer
to resume the Vigil where it stopped. The Vigil can span several sessions.
6. **Unknown is fine.** If the user does not know or prefers to answer later, write a request in
`questions/` (skill `ask-question`, `blocking: no`), keep the mark, and move on.
7. **Sensitive answers are read back** before saving (health, people's opinions, money, anything private):
"I'll write this in people/sam.md: '...'. OK?" Nothing secret is ever written: tokens and passwords go in
`.env` by the user's own hand; you only check the file exists.
8. Keep it light: short sentences, a little ceremony at chapter ends ("Chapter 2 sealed."), never a wall of
questions.
## The six chapters
| Chapter | Ask about | Write in |
| --- | --- | --- |
| 1. Names | Their language. The Seneschal's name (default "Lucan", after Sir Lucan the Butler; suggest 2 or 3 others) and how to address the user. Optionally a personality for the Seneschal (`personalities/`). | `korrigans.toml` (`language`), `agents/seneschal/IDENTITY.md` in the workspace (copy the core file with `korrigans which agents/seneschal/IDENTITY.md`, change `display_name`, `personality`) |
| 2. You and your team | Their role, field, place of work; the people they work with (role, how they relate to each project). Agents never contact these people: say so. | the user's person note (`user` in `korrigans.toml`), `people/<name>.md` from `templates/person.md`; `team: solo` in the user's note if alone |
| 3. Projects | Each project: one-line summary, status (active, paused, watch, finished), intensity (focus, background, watch), due date, people involved. Offer a Commander for focus projects (copy `agents/commanders/_template/`, Round Table name). | `projects/<p>/project.md` from `templates/project.md`, `projects/<p>/decisions.md`, `agents/commanders/<p>/` |
| 4. Needs and audience | What they want delegated first, what must never be done for them, who reads their work (advisor, clients, reviewers), how they like reports (length, time, language). | `USER.md` (Always, Never, Prefer, dated), `rules/house-rules.md`, `memory/global/preferences.md` |
| 5. Code and data | For each project: repositories and data folders, and what agents may do there (read, write on task branches, nothing). Where tools and important folders are. | `memory/global/repos-data.md` (one row per repo or data folder), `memory/global/registry.md` |
| 6. Tools and integrations | Confirm the detected harnesses and subscriptions; which modules to enable (`quota`, `todoist`, `mail`, `research`) and their settings; whether to install the Morning Chronicle timer. | `memory/global/detected.md` (confirm or correct each line), `korrigans.toml` (`modules`, module tables), module rules files such as `rules/subscriptions.toml` |
After chapter 6, run `korrigans generate` (modules and house rules change `AGENTS.md`), then `korrigans check`.
## The dubbing
When `korrigans context` reaches 100%:
1. Tell the user, in one short paragraph, what the household now knows and what it will do tonight.
2. Read them the Pact of the Moor in three lines (see `docs/lore.md` in the core): the korrigans work by
agreement, they ask rather than guess, and nothing is sealed without the user.
3. Move `BOOTSTRAP.md` to `memory/vigil.md`, add the date, commit "vigil: the Seneschal is dubbed".
4. Suggest one first real task, small enough to finish tonight.From piromagnus/korrigans-core, AGPL-3.0.
More from piromagnus/korrigans-core
- 0.561score
improve-system
piromagnus
Propose, create or install an extension (skill, script, MCP server, plugin) when a need keeps coming back.
SkillAFreshness: AClaude Code +9 - 0.561score
sort-mail
piromagnus
Sort the mail received since the last run and write the mail summary for the Seneschal.
SkillAFreshness: AClaude Code +9 - 0.561score
maintain-resources
piromagnus
Maintain resource notes and compact project.md files.
SkillAFreshness: AClaude Code +9 - 0.559score
todoist-sync
piromagnus
Build and keep the household's Todoist project, next to the user's own projects, with an owner (the user or an agent) for each delegated task.
SkillAFreshness: AClaude Code +9 - 0.559score
adjust-agent
piromagnus
Rename an agent, change its personality or its voice, add a role rule, or create the Commander of a project.
SkillAFreshness: AClaude Code +9 - 0.556score
round-table
piromagnus
The weekly review of the household - the Chronicler reads a week of journal, attempts and lessons and proposes at most three rule changes; each Commander…
SkillAFreshness: AClaude Code +9 - 0.556score
capture-idea
piromagnus
Turn an idea of the user (message, mail, passing remark) into an idea note in ideas/, never into a task.
SkillAFreshness: AClaude Code +9 - 0.556score
morning-chronicle
piromagnus
Write the user's Morning Chronicle, dated and frozen - today's board, tasks, mail, requests and quota to use.
SkillAFreshness: AClaude Code +9 - 0.556score
find-sources
piromagnus
Find sources for a charter or a request, starting from the user's library, then the web if allowed, and record every source used with its date and link.
SkillAFreshness: AClaude Code +9 - 0.556score
propagate-answers
piromagnus
Carry the user's answers to open requests into the household's memory, then close the requests.
SkillAFreshness: AClaude Code +9 - 0.556score
propose-backlog
piromagnus
When a subscription is underused before its reset, put forward one to three backlog ideas or tasks that fit in the remaining quota.
SkillAFreshness: AClaude Code +9 - 0.556score
prepare-evidence-pack
piromagnus
Prepare the evidence pack of a robust idea so the user can decide quickly whether to seal it into their knowledge base.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 20, 2026 (16 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- AGPL-3.0
- Repository
- github.com/piromagnus/korrigans-core
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from piromagnus
- 0.561score
improve-system
piromagnus
Propose, create or install an extension (skill, script, MCP server, plugin) when a need keeps coming back.
SkillAFreshness: AClaude Code +9 - 0.561score
maintain-resources
piromagnus
Maintain resource notes and compact project.md files.
SkillAFreshness: AClaude Code +9 - 0.561score
sort-mail
piromagnus
Sort the mail received since the last run and write the mail summary for the Seneschal.
SkillAFreshness: AClaude Code +9 - 0.559score
adjust-agent
piromagnus
Rename an agent, change its personality or its voice, add a role rule, or create the Commander of a project.
SkillAFreshness: AClaude Code +9 - 0.559score
todoist-sync
piromagnus
Build and keep the household's Todoist project, next to the user's own projects, with an owner (the user or an agent) for each delegated task.
SkillAFreshness: AClaude Code +9 - 0.556score
find-sources
piromagnus
Find sources for a charter or a request, starting from the user's library, then the web if allowed, and record every source used with its date and link.
SkillAFreshness: AClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get vigil running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
vigil skill. Agent Mart. https://agentmart.dev/skills/piromagnus-vigil (last updated 2026-10-24)