setup-anchor

Define or update verification anchors for recorded decisions, logs, and materials through user dialogue.

3|1|Updated Dec 9, 2025
One-click install
npx skills add https://github.com/isizono/calm --skill setup-anchor-isizono
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: setup-anchor
Source: https://github.com/isizono/calm/tree/main/skills/setup-anchor
Command: npx skills add https://github.com/isizono/calm --skill setup-anchor-isizono

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When agreements and decisions are recorded across sessions, there is often no clear way to verify later whether they are still correct. This Skill establishes an anchor for each agreement—a concrete pointer to where correctness can be checked—so future sessions can validate decisions against code, specs, or source materials. ## Core Features & Use Cases - Anchor Creation: Interactively define new verification anchors for decisions, logs, or materials that lack one. - Anchor Updates: Refresh existing anchors when code paths change, URLs move, or related decisions are replaced. - Four Anchor Types: Supports implemented code references, external spec URLs, unimplemented agreements tied to surrounding code, and fact-investigation materials. - Use Case: After a context-recomposition pass flags agreements without verifiable sources, run this Skill to walk through each one with the user and confirm concrete verification targets like src/services/checkin_tier_service.py:collect_and_assemble. ## Quick Start Ask the AI to set up anchors for the listed decisions, saying something like "create anchors for these agreements" or invoke /setup-anchor directly.

Frequently Asked Questions about setup-anchor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a verification anchor for a recorded decision?▼

Invoke the skill with a list of agreements containing entity_type, entity_id, and text. It walks you through each item interactively, proposes or asks for an anchor candidate, confirms the anchor type, and records a concrete verification target after your approval.

What anchor types are supported for agreement verification?▼

Four types are supported: implemented code references with file and function paths, external specification URLs, unimplemented agreements tied to surrounding code plus related decisions, and fact-investigation anchors pointing to stored materials.

Can I update an existing anchor when code paths change?▼

Yes. Provide the existing_anchor field in the input to trigger update mode. The skill shows the old-to-new difference and asks for your approval before finalizing the updated verification target.

Does setup-anchor write anchors to the database?▼

No. The skill has no side effects and performs no database writes. It returns an anchor mapping table with created or updated modes, and the calling agent or user handles persistence, such as material pins.

When should I not use interactive anchor setup?▼

Do not use it to verify agreements against their anchors; that validation is the responsibility of the recompose-context side. This skill only creates or updates anchors, and it avoids statically referencing the latest decision as an anchor.