#208208th of 324 in Skills, plugins and prompts
haft
Project memory and decision records for AI coding agents, via MCP and skills
- Stars
- 1.4k
- License
- custom license
- Last commit
- Sep 2026
- Last release
- Aug 2026
- Language
- Go
1,232 on 3 Apr 20261,396 on 30 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Haft is a local Go CLI and MCP server (`haft serve`) that gives coding agents a typed project memory: problems, compared options, human decisions, evidence and stale-claim tracking. It builds on the First Principles Framework, bundling the versioned FPF source and installing skills such as /h-frame, /h-compare and /h-decide into Claude Code, Codex and other hosts. State lives in `.haft/` markdown plus a per-project SQLite database.
Who it is for: Engineers using Claude Code or Codex who want auditable decision history
Strengths
- MCP server validates required fields, parity gaps and missing evidence server-side, not only in prompts
- `haft init` wires MCP config, skills and instruction sections for Claude Code and Codex
- Local-first: markdown in `.haft/` and SQLite under `~/.haft/projects/`, no hosted service
- Flags stale evidence and spec drift; binding decisions require the human
Weaknesses
- Requires learning FPF concepts; the README itself calls FPF complex
- Only Claude Code and Codex are stable; other hosts are experimental or legacy
- License is NOASSERTION in detected metadata, so terms need checking
- Schema migrations and carrier semantics add operational overhead on upgrade
What it needs
- no GPU
- Needs SQLite, Claude Code, Codex or another MCP host, haft-embed sidecar (optional)
Quick answers
Is haft open source?
Its code is public on GitHub under its own license terms; read them before you use it at work.
Also in Skills, plugins and prompts
See all 324| Rank | Project | Score |
|---|---|---|
| 1 | impeccableDesign skill with 24 slash commands for AI coding agents' frontend output | 96 out of 100 |
| 2 | skillsAgent skills for planning, TDD, debugging and code review | 95 out of 100 |
| 3 | eccAgents, skills, hooks and rules bundle for Claude Code, Codex and other harnesses | 95 out of 100 |
| #4 | gstackMarkdown slash-command skills that give Claude Code planning, review, QA and ship roles | 95 out of 100 |
| #5 | cavemanCuts token use in coding agents with terse replies and compressed inputs | 95 out of 100 |