pythia

Oracle's MCP gives your agent a connection. pythia gives it the judgment to use it.

English · Tiếng Việt

An Agent Skills + CLI kit for developing PL/SQL on Oracle Database with AI coding agents (Claude Code, Codex, Cursor — any of the 76 agents npx skills supports). Explore schemas too big to dump, measure blast radius before touching anything, and land changes through a snapshot-verified write path that never lies about rollback.

Why ask the database instead of reading dumps

A real mid-size system, repo export vs live database, audited in 2026:

Code that "reads fine" against the dump references types and packages the dump never heard of. Every pythia command asks the live data dictionary instead — and every truncated output says so, so an agent never mistakes a partial answer for a full one.

How it works

developer chats with the agent
        │
skills/  teach the agent when to ask, when to stop, when to ask YOU
        │
pythia   CLI — expert queries, impact analysis, the six-step write path
        │
Oracle   data dictionary: ALL_SOURCE, ALL_DEPENDENCIES, ALL_ERRORS, PL/Scope

The write path is the heart: snapshot → impact → preview → apply → verify → report. DDL self-commits in Oracle — the snapshot is the only real undo, so it always runs first and no flag can turn it off. A 6-hex token binds the write to exactly what was previewed; exit codes make honesty machine-readable (0 clean · 1 refused · 3 written but broken — never reported as success).

Install

npx pythia-plsql           # everything: pip install + skills picker + config scaffold

Or the same thing piecewise:

pip install pythia-plsql   # the CLI (thin driver — no Oracle Instant Client needed)
pythia install             # skills into your agent + .pythia/connections.json scaffold
pythia check               # fill in connections.json first, then verify

The pip package is the whole kit: with Node.js present, pythia install runs npx skills add (77 agents, symlinked updates; --source <git-url> for internal mirrors) — without Node it copies the bundled skill pack into .agents/skills/, the universal layout Claude Code, Codex and Cursor all read (one copy, so no skill shows up twice). Skills alone: npx skills add thaildhe172591/pythia, or /plugin marketplace add thaildhe172591/pythia.

pip install is once per machine; pythia install is once per project — run it in each repo's root to drop the skills and a fresh .pythia/connections.json there. The CLI always reads the config of the project you are standing in (searched upward from the current directory, no global fallback), so one global CLI never mixes up databases.

Running from a clone works too — python scripts/pythia.py <command>; every printed follow-up command matches how you invoked it. Windows, macOS, Linux and WSL are all CI-tested.

Commands

Every command takes --json (machine output), --conn (pick a connection), and caps output with explicit truncation markers so context windows stay intact.

Your house style is config, not folklore: put naming patterns in .pythia/conventions.json and apply previews warn when a new object's name drifts; put the prose rules in .pythia/conventions.md and the skills make every agent read them first (pythia conventions shows both).

Security & write policy

The account is the real security layer — the policy file is an application-side fence. Give the agent its own revocable credential with proxy authentication (agent_user[schema_owner], no ANY privileges, no owner password shared):

pythia agent-user --save   # SQL for the DBA + matching credential saved as <conn>_agent
pythia check               # after the DBA ran it: proxy session, warning gone

One run does both — the password is regenerated each run, so the SQL and the saved config must come from the same run. Optional convenience: doing it by hand with examples/agent-user-setup.example.sql works just as well. pythia check warns when the session runs with more power than the task needs.

Per-group write policy, .pythia/policy.json (defaults shown):

The groups that cannot be snapshotted default to deny — and the refusal says that, instead of "policy forbids it". Anonymous PL/SQL blocks are refused outright. Unrecognized statements are refused, never guessed into a group.

Reads may flow through Oracle's official SQLcl MCP server (sql -mcp, keep -R 4); it audits every interaction in DBTOOLS$MCP_LOG. Writes never do — only pythia apply has the snapshot, preview, verify and journal.

Skills

Seven skills teach the agent the workflow — superpowers-style gates, not suggestions:

pythia-setup · pythia-explore · pythia-impact (impact before any change) · pythia-write (copy the codebase's conventions) · pythia-apply (the gate: the developer sees the preview and approves in chat before anything is written) · pythia-review (seven antipatterns) · pythia-skill-author (capture your team's workflow as a new skill, mined from the live schema).

Compatibility

Star History

Contributing

Tests need no database — the fakes prove the safety properties (snapshot before write, deny touches nothing, stale tokens refused). See CONTRIBUTING.md.

MIT — see LICENSE.