AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified Apache-2.0 Self-run

Map Init

skill-allemaar-open-skills-map-init · by allemaar

Onboard Map Your Knowledge into a user's agent directives — check whether it is already installed, or offer to add the routing rules globally or per-project, for EVERY agent platform on the machine, with explicit consent and a per-platform manual fallback. Trigger when the user says "install map your knowledge", "set up myk", "make this my agents' default", or "/map-init". Idempotent and consent-…

No reviews yet
0 installs
0 views
view→install

Install

$ agentstack add skill-allemaar-open-skills-map-init

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-allemaar-open-skills-map-init)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
9d ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Map Init? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

/map-init

The onboarding door for Map Your Knowledge (MYK). It does exactly one job: make MYK the default way of working for the user's agents — with their explicit consent, on every agent platform they use, with every outcome reported exactly.

> kernel-version: MYK v2.3 f2f96f2de49b4863bca55ee8f6004d24e00574a7db5e7e5ef0e3cb28c42510cf

Why every platform matters (say this to the user)

A knowledge graph is maintained by every agent that touches the vault. If one agent carries the rules and another does not, the uninformed one can undo in a session what the others maintain carefully. MYK works as a modus operandi only when EVERY resident agent loads it — that is why this skill offers the installation for all detected platforms, not just the one running it.

Steps

  1. Detect platforms. Identify every agent platform configured on this machine by its directive surface — e.g. Claude (~/.claude/CLAUDE.md, project CLAUDE.md/.claude/), Codex (~/.codex/AGENTS.md, project AGENTS.md), and any other agent instruction file the user names. Read-only detection; never guess at unfamiliar formats — ask.
  2. Check idempotently. For each platform, check whether the FULL payload is already present — all three elements: the load-the-rules routing clause, the orient-from-the-root-map clause, AND the born-mapped clause (a mere map-rules name reference is NOT installed; partial presence = offer to complete the missing clauses, shown as a diff). If the full payload is present everywhere: report "Nothing to do — you already have Map Your Knowledge installed" per platform, and stop.
  3. Offer, with the choice and the benefits. For each platform missing the rules, present: (a) the choice — global (all projects) or project-local (this project's directive file only); (b) the exact text to be added (small, shown in full — see the payload below); (c) the benefits in plain words: agents orient from the scope's map before acting, new data is born mapped, existing structures are respected via the elastic verdicts, and every agent behaves consistently. This is optional and the user decides — per platform.
  4. Apply exactly what was consented. Insert the payload into the consented file(s) only. Show the diff before writing. Pre-check every consented target is writable BEFORE the first write; each file write is atomic; if a later platform's write still fails, report precisely which platforms succeeded and which did not, and offer the paste-fallback for the failed ones — the user is never left uninformed about a mixed state.
  5. Fallback: instruct, never force. Where automation cannot edit a platform's directives (unknown format, no file access, a hosted agent like a web UI), print the exact snippet plus per-platform instructions: where the file lives, where to paste, and how to verify (ask the agent "what do you do before organizing vault markdown?" — the answer should name the map rules).

The payload (what actually gets injected)

Two sentences, adapted to the platform's convention:

- [myk.routing] Before creating, editing, moving, renaming, classifying, mapping,
  rolling up, archiving, or evaluating the organization of vault or project
  Markdown, load the map-rules skill and obey it; use map-this for bounded
  organize/housekeep work. When entering an organized scope, orient from its
  declared root map before acting; new files inside an organized scope are
  born mapped (creation and mapping are one approved set).

Plus, where the platform supports skills, a pointer to install the map-* family if absent.

Safety invariants

  • Idempotent: running twice changes nothing the second time.
  • Consent-first, per platform, with the exact diff shown; global-vs-project is the user's call.
  • Per-file atomic and mixed-state honest: every write is atomic per file, and the user always receives exact per-platform outcomes — never an unreported mixed state.
  • Never touches vault content, other directives, or any file beyond the consented directive files.
  • The user's existing directives are house style: if their file has its own structure/conventions, the insertion adapts to it (placement, list style) — the elasticity rules apply to THIS file too.

> Human output. This skill's handler-facing output obeys the human-output > contract (human-output/SKILL.md).

> Self-improvement. On completion, run the Self-Improvement Protocol (self-improve/SKILL.md): if this run surfaced a concrete, blocking-or-recurring weakness in this skill, propose a specific fix for the handler to approve. Conservative — silent otherwise. Never auto-apply.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.