LockedIn CLAUDE.md

Project instructions for LockedIn, a Claude Code plugin that turns a user's stored experiences into structured notes and documents such as résumés and cover letters.

In plain words
What is it for?
Use it when changing the plugin, its skills, command-line helpers, or document-generation logic, while preserving the separate input vault and generated-output folders.
Why use it?
It explains where the plugin code belongs and where the user's private data and generated documents are kept. This separation helps prevent personal information from being mixed into the source repository.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/daypunk/lockedin/claude-md
Clone the repo
git clone --depth 1 https://github.com/daypunk/LockedIn
Per session 2,857 This file is loaded in full into every session.
When invoked 2,857 The same file — it is already loaded in full.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.02857 $0.02857
Opus 5 $0.01429 $0.01429
Sonnet 5 $0.00571 $0.00571
Haiku 4.5 $0.00286 $0.00286

Measured yesterday against content hash 703ab352fa70, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

LockedIn CLAUDE.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

- **Never auto-modify `~/.claude/settings.json`** outside the
CLAUDE.md · 209 lines

How it starts

The opening of the file, as written. The whole thing — 209 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Project-level instructions, loaded automatically by Claude Code when working in this repo. Read this first if you are picking up the project in a fresh session. The file is designed to be self-contained enough that you can resume without prior conversation context.

What this is

LockedIn is a Claude Code plugin that organizes a user's experience (career, meetings, projects, learning, decisions) into a typed markdown ontology they own at ~/Documents/LockedIn/, then renders text artifacts (English resume, Korean cover letter) from the same vault. Two stores by design.

  • Store A is the user's input, structured into the vault under ~/Documents/LockedIn/experience/.
  • Store B is ~/Documents/LockedIn/outputs/, the artifacts rendered from A.

This repo holds the engine. It contains the plugin manifests, skill files, CLI helpers, and tests. Users' actual data lives at ~/Documents/LockedIn/, never in this repo.

Tagline: Personal experience knowledge graph for Claude Code. Zero learning curve. Distribution is the Claude Code plugin marketplace (/plugin marketplace add daypunk/LockedIn).

The lowercase identifier lockedin is used everywhere a machine parses it (Python package, plugin name, slash command, env vars). The brand display LockedIn is used in user-facing copy (README headers, descriptions, sentences in prose).

Current state

Layer State
Plugin marketplace manifests at .claude-plugin/marketplace.json and plugins/lockedin/.claude-plugin/plugin.json
One-time setup wizard /lockedin:setup at plugins/lockedin/commands/setup.md (HUD wiring, Q&A language, vault path)
Ontology v3. 15 entity types, 15 edge predicates, JSON Resume / Schema.org / FOAF aligned. Per-type field contracts plus aliases (person, company) and provenance (every type). Edge domain and range enforced by lockedin validate. See lockedin/ontology/schema.py.
Render skills lockedin (main), lockedin-render-jaso (calibrated, 5 pass + 5 fail fixtures), lockedin-render-resume-en (calibrated, 3 personas). All under plugins/lockedin/skills/.
HUD lockedin X.Y.Z │ 5h:NN% · wk:NN% │ experience: Nn · Me. Pulls real utilization from Anthropic OAuth API, falls back to vault-only display when OAuth is unavailable. Same script in two places (plugins/lockedin/scripts/hud.py standalone, lockedin/commands/hud.py package). The standalone defers to the package when available.
Deterministic CLI install (with --setup-hud, --remove-hud, lifecycle), init --fixture FILE, ingest --dry-run, validate, migrate, experience, doctor, template, hud.
Skill-only commands init (interactive), ingest (smart), render jaso/resume, query. Typed in plain shell, the CLI prints a redirect (exit 3) and points at Claude Code.
OSS infrastructure CONTRIBUTING.md, CODE_OF_CONDUCT.md (Contributor Covenant 2.1), .github/ISSUE_TEMPLATE/{bug_report, feature_request, korean_reviewer_onboarding, config}.yml.
Tests 200+ passing across import, ontology, storage round-trip, install lifecycle, doctor, validate v3, init from fixture, template, ingest dry-run, ingest format parsers (pdf/docx/markdown/text), hud, interview resumability, render-interview and render-ideas calibration, README-promise lint.

Read the full file on GitHub · 209 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 209 lines · 2,857 tokens per session scan B 703ab352fa70

Subscribe to this mod's changes

LockedIn CLAUDE.md is an instructions file published in the GitHub repository daypunk/LockedIn (127 stars, last pushed 3mo ago), licensed MIT. It adds 2,857 tokens to every session, about $0.0143 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.