How do I install the Data Store skill?

Run npx skillmds add majiayu000/data-store 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 Data Store safe to use?

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

Which AI agents work with Data Store?

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 Data Store free to use?

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

Who published Data Store?

majiayu000 (@majiayu000) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Data Store by majiayu000 — Agent Skill

Name: Data Store
Rating: 4.666666666666667 (3 reviews)
Author: majiayu000

File contents

Agent Data Store Convention

Defines how agents store and access persistent project-level data across sessions.

Directory Layout

.epost-data/              ← gitignored, project-level runtime state
  {domain}/
    {domain}.json         ← primary DB (required)
    artifacts/            ← generated patches, diffs, temp files (optional)

a11y Domain Layout

.epost-data/a11y/
├── fixes/
│   ├── findings/         ← audit-a11y JSON reports (audit-YYMMDD-HHMM.json)
│   ├── patches/          ← fix-a11y unified diffs (finding-{id}-YYMMDD.diff)
│   └── reviews/          ← review-a11y JSON reports (review-YYMMDD-HHMM.json)
├── README.md             ← structure guide
├── analysis.md           ← trend summary
└── known-findings.json   ← findings DB

Rules

Rule	Detail
Always `(if exists)`	Never hard-fail on missing file — gracefully absent on first run
Gitignored	Runtime state, not source-tracked; must be in `.gitignore`
One agent = one domain	Agent owns its domain exclusively; never reads another agent's
Schema in source	Schema lives at `packages/{pkg}/assets/{domain}-schema.json`, tracked in git

Agent Declaration Template

Paste into the agent's system prompt body under a Data Store section:

## Data Store
- **DB:** `.epost-data/{domain}/{domain}.json` (if exists)
- **Artifacts:** `.epost-data/{domain}/artifacts/` (if exists)
- **Schema:** `.claude/assets/{domain}-schema.json`

Domain Registry

Domain	Owner Agent	Primary File	Package
`a11y`	epost-a11y-specialist	`known-findings.json`	packages/a11y

Add a row here when introducing a new agent data domain.

Schema Conventions

Minimum required fields for any {domain}.json:

{
  "version": "1.0",
  "created_date": "YYYY-MM-DD",
  "last_reviewed_date": "YYYY-MM-DD",
  "entries": [
    {
      "id": 1,
      "resolved": false,
      "resolved_date": null
    }
  ]
}

Bump version string on schema changes (new fields = minor, breaking = major)
All commands check resolved before operating on an entry
last_reviewed_date updated any time the DB is written

Adopting the Pattern (New Domain)

Create schema: packages/{pkg}/assets/{domain}-schema.json
Add domain row to the registry table above
Add "Data Store" section to agent body using the template
Reference paths in commands with (if exists) guards
Add .epost-data/ to project .gitignore (if not already present)

---name: data-storedescription: (ePost) Use when an agent needs to persist or retrieve project-level data across sessions via .epost-data/--- # Agent Data Store Convention Defines how agents store and access persistent project-level data across sessions. ## Directory Layout ``` .epost-data/ ← gitignored, project-level runtime state {domain}/ {domain}.json ← primary DB (required) artifacts/ ← generated patches, diffs, temp files (optional) ``` ### a11y Domain Layout ``` .epost-data/a11y/ ├── fixes/ │ ├── findings/ ← audit-a11y JSON reports (audit-YYMMDD-HHMM.json) │ ├── patches/ ← fix-a11y unified diffs (finding-{id}-YYMMDD.diff) │ └── reviews/ ← review-a11y JSON reports (review-YYMMDD-HHMM.json) ├── README.md ← structure guide ├── analysis.md ← trend summary └── known-findings.json ← findings DB ``` ## Rules | Rule | Detail | |------|--------| | **Always `(if exists)`** | Never hard-fail on missing file — gracefully absent on first run | | **Gitignored** | Runtime state, not source-tracked; must be in `.gitignore` | | **One agent = one domain** | Agent owns its domain exclusively; never reads another agent's | | **Schema in source** | Schema lives at `packages/{pkg}/assets/{domain}-schema.json`, tracked in git | ## Agent Declaration Template Paste into the agent's system prompt body under a **Data Store** section: ```markdown ## Data Store - **DB:** `.epost-data/{domain}/{domain}.json` (if exists) - **Artifacts:** `.epost-data/{domain}/artifacts/` (if exists) - **Schema:** `.claude/assets/{domain}-schema.json` ``` ## Domain Registry | Domain | Owner Agent | Primary File | Package | |--------|-----------------------|-----------------------|------------------| | `a11y` | epost-a11y-specialist | `known-findings.json` | packages/a11y | *Add a row here when introducing a new agent data domain.* ## Schema Conventions Minimum required fields for any `{domain}.json`: ```json { "version": "1.0", "created_date": "YYYY-MM-DD", "last_reviewed_date": "YYYY-MM-DD", "entries": [ { "id": 1, "resolved": false, "resolved_date": null } ] } ``` - Bump `version` string on schema changes (new fields = minor, breaking = major) - All commands check `resolved` before operating on an entry - `last_reviewed_date` updated any time the DB is written ## Adopting the Pattern (New Domain) 1. Create schema: `packages/{pkg}/assets/{domain}-schema.json` 2. Add domain row to the registry table above 3. Add "Data Store" section to agent body using the template 4. Reference paths in commands with `(if exists)` guards 5. Add `.epost-data/` to project `.gitignore` (if not already present)

Frequently asked questions