v1.5.0

Postgres Introspection Skill v1.5.0

This release adds a purpose document for developers and a tracked home for local template feedback.

Highlights

• docs/PURPOSE.md explains why the skill exists and what it gives you as a developer, with a section written for newcomers. It is linked from the README and sits beside it as a front-door document.
• A tracked .skill-template-feedback/ folder, with its own README and a .gitkeep, is the local home for notes about the skill template. Its contents are git-ignored, so working notes stay local while the convention travels with the repository.

Why it matters

A new developer can now read one short document to understand what the skill is for and why it helps, before any technical detail. The feedback folder gives maintainers a consistent place to capture template gaps and route them upstream.

Notes

Release tag: v1.5.0.

v1.4.0

Postgres Introspection Skill v1.4.0

This release makes the front page welcoming and rounds out the plugin metadata.

Highlights

• The README now reads as a front page. It opens with what the skill gives you in plain language, then how to install and use it, and moves the conceptual and internal detail into the docs.
• Both plugin manifests now declare "license": "MIT", and the Claude plugin manifest gains a top-level "displayName", after a field-by-field check against the Claude Code and Codex plugin specifications.

Why it matters

A newcomer landing on the repository now sees the value and the install path first, instead of internal structure. The plugin packages carry complete, spec-aligned metadata for the install surfaces that display it.

Notes

Release tag: v1.4.0.

v1.3.0

Postgres Introspection Skill v1.3.0

This release corrects the standalone skill install locations to match the Agent Skills standard, so the documented paths actually work across tools.

Highlights

• .agents/skills/ is now the primary documented location for the standalone skill. It is the cross-tool open standard, recognized by Codex and GitHub Copilot, and scanned up to the repository root.
• The incorrect .codex/skills/ path was removed; Codex reads .agents/skills/.
• Per-tool guidance is clearer: .claude/skills/ (and ~/.claude/skills/ for personal use) for Claude Code, and .github/skills/ for GitHub Copilot.

Why it matters

A user following the old instructions would have dropped the skill into .codex/skills/, where Codex does not look. The corrected paths align with the current Agent Skills standard and each tool's documentation.

Notes

Release tag: v1.3.0.

v1.2.0

Postgres Introspection Skill v1.2.0

This release makes installation obvious for normal use. The README now leads with downloading a release ZIP and installing it, instead of reading like a clone-and-build project.

Highlights

• The README has an Install section near the top: download the latest release from the Releases page and install the asset for your tool, the standalone skill, the Claude Code plugin, or the Codex plugin.
• The maintainer commands (npm run validate and npm run package) moved into a "For maintainers and contributors" section, so end users are not told to clone and build.
• docs/INSTALL.md now points to the Releases page and names the ZIP asset for each install option.

Why it matters

A person who just wants to use the skill can now get it in a few clicks from the Releases page and follow per-tool install steps, without touching Node or the repository source.

Notes

Release tag: v1.2.0.

v1.1.0

Postgres Introspection Skill v1.1.0

This release documents and hardens the maintenance process. The skill's behavior is unchanged; the focus is making the release process and the repository workflow clear and unambiguous for the humans and agents who maintain it.

Highlights

• docs/RELEASING.md now describes the whole process: how changes land through a pull request, what the main ruleset enforces, the CI check that gates merges, the repository safeguards, and the tag-triggered draft release.
• The release process has a single source. AGENTS.md points to docs/RELEASING.md instead of repeating its steps, so the instructions cannot drift apart.
• Release instructions in CONTRIBUTING.md and the pull request template were reconciled, and the hardcoded npm run package -- v1.0.0 was replaced with a version-neutral vX.Y.Z.

Why it matters

A maintaining agent or contributor now finds one consistent description of how to land a change and cut a release, with no conflicting or version-specific instructions across the documents.

Notes

Release tag: v1.1.0.

v1.0.0

Postgres Introspection Skill v1.0.0

First release of the postgres-introspection skill, which teaches an AI coding agent to build a database introspection capability inside the repository it works in.

Highlights

• A read-only introspection tool the agent recreates in the target repository, fitted to its stack, that keeps working without the skill.
• A runnable PostgreSQL reference implementation covering the database, the cluster, a connection pooler, and a REST layer.
• The method and its invariants, the classification path, and the minimal-greppable rendering discipline.
• A commenting discipline that captures the architect's intent as compact database comments, with the database as the single source of truth and documentation deriving from it.
• Guidance for reading the state and using the reports with judgment, including the alignment loop and conflict handling.
• Data-safety rules that protect existing data and database operability and require informed consent for anything consequential.

Why it matters

The real state of a database lives in the database, and the reasons behind its shape usually live nowhere durable. This skill gives a repository a tool to export the current state into shareable, credential-free, version-controlled files, and a discipline to keep the intent next to the data. Together they let teams and different AI agents reason about the database safely and see how it evolves over time.

Notes

Release tag: v1.0.0.