How do I install the Vestige Context skill?

Run npx skillmds@latest add conorluddy/vestige-context in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Vestige Context safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Vestige Context?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Vestige Context free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Vestige Context?

conorluddy (@conorluddy) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Vestige Context by conorluddy — Agent Skill

Name: Vestige Context
Author: conorluddy

Vestige Context

Render the project's Vestige context pack — summary, recent decisions, open questions, recent activity — as a budget-bounded brief. Use at session start, before unfamiliar work, after a long pause, or when the user says "what's the state of this project?", "where are we?", "catch me up", "what have we decided?", "what's outstanding?", "give me the rundown". Returns a JSON envelope with `summary`, `decisions[]`, `open_questions[]`, `recent[]`. Token budget defaults to 1200 — pass `--budget-tokens` to adjust.

conorluddy d36da6e 3 files · 5.0 KB Updated Sep 17, 2026

File contents

Get the project context pack

Pull the durable Vestige memory an agent should read before doing serious work in this project. Same shape MCP tooling exposes via vestige_get_project_context.

When to fire

Session start. First time you're working on this project today, fire this once.
Cold start on unfamiliar code. About to refactor a module you haven't touched? Pull the pack first.
User explicitly asks. "What's the state?", "remind me where we are", "summarise this project for me".
Before a non-trivial decision. If you're about to commit to architecture, the pack tells you what you're already committed to.

Tie-breaker vs vestige-recall: the pack is broad (project-wide). Recall is narrow (one query). If you have a specific question, use recall — it's cheaper.

How to invoke

vestige context --json

--budget-tokens <n> (default 1200): caps the total pack size. Lower it (300–500) when context is tight; raise it (2400+) for a deep dive.
--per-section <n> (default 8): caps each list section.
--json (recommended for agents): structured envelope. Skip for human-readable markdown.

After invocation

The JSON envelope:

{
  "project_name": "…",
  "summary": { "id": "mem_…", "one_liner": "…", "available_depths": […] } | null,
  "decisions": [{ "id": "mem_…", "one_liner": "…", "score": …, "type": "decision" }, …],
  "open_questions": [ … ],
  "recent": [ … ],
  "warnings": []
}

For any card whose one_liner isn't enough, use vestige-show <id> to expand at higher fidelity. Cite handles (mem_…) when you reference a finding.

Idempotence & dedup

Read-only — calling repeatedly is safe and cheap, but pointless. Skip if:

You already pulled the pack this session and nothing has changed.
The user just asked a narrow question — don't pull the whole pack to answer "what's our SQLite version?", use vestige-recall "sqlite" instead.

---name: vestige-contextdescription: Render the project's Vestige context pack — summary, recent decisions, open questions, recent activity — as a budget-bounded brief. Use at session start, before unfamiliar work, after a long pause, or when the user says "what's the state of this project?", "where are we?", "catch me up", "what have we decided?", "what's outstanding?", "give me the rundown". Returns a JSON envelope with `summary`, `decisions[]`, `open_questions[]`, `recent[]`. Token budget defaults to 1200 — pass `--budget-tokens` to adjust.---# Get the project context packPull the durable Vestige memory an agent should read before doing serious work in this project. Same shape MCP tooling exposes via `vestige_get_project_context`.## When to fire- **Session start.** First time you're working on this project today, fire this once.- **Cold start on unfamiliar code.** About to refactor a module you haven't touched? Pull the pack first.- **User explicitly asks.** "What's the state?", "remind me where we are", "summarise this project for me".- **Before a non-trivial decision.** If you're about to commit to architecture, the pack tells you what you're already committed to.Tie-breaker vs `vestige-recall`: the pack is *broad* (project-wide). Recall is *narrow* (one query). If you have a specific question, use recall — it's cheaper.## How to invoke```bashvestige context --json```- **`--budget-tokens <n>`** (default 1200): caps the total pack size. Lower it (300–500) when context is tight; raise it (2400+) for a deep dive.- **`--per-section <n>`** (default 8): caps each list section.- **`--json`** (recommended for agents): structured envelope. Skip for human-readable markdown.## After invocationThe JSON envelope:```json{ "project_name": "…", "summary": { "id": "mem_…", "one_liner": "…", "available_depths": […] } | null, "decisions": [{ "id": "mem_…", "one_liner": "…", "score": …, "type": "decision" }, …], "open_questions": [ … ], "recent": [ … ], "warnings": []}```For any card whose `one_liner` isn't enough, use `vestige-show <id>` to expand at higher fidelity. Cite handles (`mem_…`) when you reference a finding.## Idempotence & dedupRead-only — calling repeatedly is safe and cheap, but pointless. Skip if:- You already pulled the pack this session and nothing has changed.- The user just asked a narrow question — don't pull the whole pack to answer "what's our SQLite version?", use `vestige-recall "sqlite"` instead.

Frequently asked questions