Installable agent skills that teach an AI coding agent
to use PgBeam for safe Postgres access. Each skill is a
SKILL.md with YAML frontmatter (name, description) and a body the agent
reads when a task matches.
| Skill | What it teaches |
|---|---|
pgbeam-connect |
Wire an AI agent to Postgres safely: get a scoped credential, choose a guarded connection string or the hosted MCP endpoint, and paste-ready Claude Code, Cursor, and VS Code config. |
pgbeam-policy |
Author a policy profile as code: access mode, table allow and deny lists, PII masking, row filters, budgets, write mode, with CLI and Terraform. |
pgbeam-mcp-usage |
Drive the hosted Postgres MCP tools well once connected: prefer schema_catalog, read the policy errors, work within read-only and masking. |
With the open skills tool:
npx skills add pgbeam-connectOr copy a SKILL.md into your agent's skills directory (for example
.claude/skills/pgbeam-connect/SKILL.md for Claude Code, or .agents/skills/
for the cross-agent standard).
Agents can also discover these skills over HTTP from the PgBeam site, which serves a machine-readable index with per-skill integrity digests:
- Index:
https://pgbeam.com/.well-known/agent-skills/index.json - Each body:
https://pgbeam.com/skills/<name>/SKILL.md