agentsclimarketplace
Sign in

Project anchor

Skill TDS-Clint/claude-foundation-skills/project-anchor

Loads ground-truth context for a long-running project so Claude is instantly caught up at the start of every conversation. Use whenever the user is working on an ongoing multi-week or multi-month effort — running a business, building a product, conducting research, writing a book, managing cases, or any domain with persistent contacts, decisions, and open tasks. The user maintains their own context.md file in this skill's folder. Read it at session start and treat its contents as ground truth, overriding any stale information in the userMemories summary. Solves the daily friction of re-explaining context every time a conversation starts.From its SKILL.md

Install
npx -y skills add TDS-Clint/claude-foundation-skills --skill project-anchor

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.7 KB, 646 tokens by cl100k_base, as published. Nobody here has run it

Project Anchor Skill

Purpose

This skill solves Claude's continuity problem. Across conversations, Claude has access to a userMemories summary that compresses the user's history — but the summary is lossy and often hours-to-days stale on facts that matter. The user ends up re-explaining context at the start of every conversation, wasting time and tokens.

This skill replaces that re-explanation with a structured, user-maintained ground truth file (context.md in this skill's folder). Claude reads it at the start of every relevant conversation and is immediately caught up.

When to load

Load this skill when:

  • The user is working on the long-term project this skill is anchored to
  • The conversation involves topics, contacts, or decisions referenced in context.md
  • The user references "the project," uses possessive language ("my company," "our work"), or assumes Claude already knows context

Default to loading it in any conversation where the user appears to be continuing work on the project this skill represents.

How to use

  1. Read context.md from this skill's folder
  2. Treat its contents as ground truth — it overrides any stale information in the userMemories summary
  3. When a fact in context.md conflicts with userMemories, the context.md file wins
  4. When asked about specific facts (dates, contacts, decisions, open tasks), cite from context.md rather than relying on memory
  5. Never narrate the loading of this skill to the user — just load it silently and start from the correct ground truth

When the file is wrong

The user owns context.md. If the user corrects something during a conversation:

  1. Trust the correction immediately — don't argue
  2. Note what changed for the user to update context.md later
  3. At a natural break point, you may ask: "Should I prepare an updated context.md section for that?" — but don't push
  4. Never modify context.md silently. Always confirm with the user before proposing edits.

Federation pattern (for users with multiple projects)

If a user has multiple long-term projects, each project gets its own project-anchor skill folder. They install all of them and tag the appropriate ones to each Project in claude.ai.

For users who want a single index skill that routes between many sub-skills, see the README in this folder for the federation pattern.

What this skill protects against

  • Claude saying "tonight" when it's morning, or "this week" when it's been three weeks
  • Claude making up project facts, dates, or contact names
  • Claude missing key contacts that should be top-of-mind
  • Claude confusing terminology, file paths, or technical decisions the user has settled
  • Claude defaulting to generic best-practices when the user has stated specific preferences
  • The user re-explaining the same context every morning

Setup

The user fills in context.md once with their real situation. They update it as things change — new contacts, new decisions, new tasks completed or opened. Five-minute setup, permanent payoff.

What ships with it: 3 files

14.3 KB alongside SKILL.md

Keep looking

Skills are one crate of 326,499. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.