SkillFoxx

PlanDB CLI

PlanDB is an issue tracker for AI agents: a dependency graph, atomic task claiming, notes and critical path analysis in SQLite.

  • 112 ★ on the repository
  • Last commit
  • Apache-2.0
CLI

Extended access

Why this level

  • The installer edits agent instruction files in the home directory.
  • The install command runs a script from the network.
All reasons and checks

agent-field/plandbAuthor:Agent-Field

Install

Manual install

curl -fsSL https://agentfield.ai/plandb/install | bash

The first command of the Install section. The script downloads the binary and configures detected agents.

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The installer edits agent instruction files in the home directory.
  • The install command runs a script from the network.

This is third-party code. Review the repository files before installing.

What it does

A local issue tracker in a single binary on SQLite, with no cloud and no accounts. An agent lays work out as a compound graph: tasks contain subtasks to any depth and dependencies link them across branches. `split` divides a stuck task into parallel ones, `insert` adds a missed step with dependencies rewired and `pivot` replaces a whole subtree. `plandb go` claims the next task atomically, so two agents cannot take the same one. `plandb context` stores discoveries and decisions, and they surface through BM25 search when a related task is claimed. There are `critical-path`, `bottlenecks`, `--pre` and `--post` conditions, YAML pattern export and a terminal dashboard `plandb watch`.

Who it is for. Developers who run several agents on one task and want a shared plan with dependencies and memory across sessions.

Good fit when

  • Several agents work in parallel on one project and must not duplicate tasks.
  • A task turned out harder and must split into subtasks with dependencies.
  • Findings from one session should surface for an agent days later.
  • You want to see the longest dependency chain and bottlenecks.

Not a fit when

  • You need a tracker for humans with shared cloud access.
  • Installing with a `curl | bash` script is not acceptable on the machine.
  • There is one agent and the task is simple and needs no plan.

Example request

/plandb build a CLI todo app in Python with add, list, complete and delete commands, store data in JSON and include tests.

Limitations

The installer edits agent instructions itself: it creates `~/.claude/rules/plandb.md` and a skill in `~/.claude/skills/plandb/`, and appends to `~/.codex/AGENTS.override.md` and `~/.gemini/GEMINI.md`. For Cursor the instructions are only printed. `curl | bash` runs a script from the network, so read it first. `--binary-only` installs just the binary. Plan state lives locally in SQLite, and shared use across machines is not stated. The last commit is from September 2026.

How to disable. Delete the `plandb` binary, the rules and skill in `~/.claude`, and the blocks appended to `AGENTS.override.md` and `GEMINI.md`. Delete SQLite plan files in projects by hand.

FAQ

How do I install PlanDB?

Run `curl -fsSL https://agentfield.ai/plandb/install | bash`. The script installs the binary and configures detected agents. From source: `cargo install --path .`.

How do I install only the binary?

Add `-s -- --binary-only` to the install command.

How do two agents avoid the same task?

`plandb go` uses atomic operations, so only one agent gets a task.

Which interfaces exist?

A CLI for shell agents, the `plandb mcp` MCP server for Claude Code, Cursor and Windsurf, and an HTTP API.

How do I watch progress?

`plandb watch` shows a live terminal dashboard.

Discussion

No questions or reviews yet.

CLI tools

A small issue tracker for AI agents: tasks, epics and dependencies live in a project SQLite file, with a local board.

CLIMinimal access88

A project tracker on SQLite in a single binary: CLI, web UI and a /pad skill through which an agent manages tasks, plans and ideas

CLIExtended access207

A SQLite issue tracker for AI agents: sessions with handoff notes, dependencies, templates and Claude Code hooks

CLIExtended access367

A task tracker for agents: tasks with context and result live in JSONL beside the code and outlive sessions

CLIMinimal access410

Local SQLite task manager for DeepSeek Harness: the agent claims and submits work, and only a human accepts it.

PluginMinimal access327

The MCP server of the local Taskdog task manager: an agent creates tasks, runs their lifecycle and splits big ones into subtasks.

MCP serverExtended access365

README in short

The README presents PlanDB as an agent tracker, gives one-command installation with a table of configured frameworks, sample prompts, explanations of the graph, context and commands and a demo of a four-task plan.

README badge

Report a problem