Lore
Mines your SpecStory coding histories into corroborated, outcome-labeled skill candidates you forge into real skills.
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 20, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 0659955
Cloned the repo and actually ran the engine (pure Node stdlib, zero npm deps): index parsed 6 multi-agent sessions across 2 nested projects, and report surfaced the corroborated pattern "build:run × go build ▸ go test ▸ golangci-lint run" (3 sessions, 2✓/1✗ outcomes labeled from the user's next reply, exact commands from executed shell blocks, path:line refs, git author attribution) plus a verbatim pass-through summary — a clear, specific win over baseline transcript-guessing. Security scan clean: only local `git` via execFileSync for authorship, no network imports, no curl|sh, and the plan-validation hook correctly denies non-engine forge plans. All referenced scripts (mine-skills.mjs, hooks/validate-plan.mjs, workflows, lib/*) exist and `forged check` idempotency works. TRIGGER — SHOULD fire: "what could I make into a skill from my sessions?", "mine my lore and forge skills from my history", "turn my last 30 days of coding into reusable skills"; SHOULD NOT: "write me a skill for deploying to Vercel" (author-from-scratch, no mining), "summarize what I did in yesterday's session" (session summary, not skill extraction) — 5/5 correct. Only friction: requires the separate SpecStory recorder to have captured sessions, but the skill+engine themselves are fully self-contained and measured working. Exceptionally honest docs (names its own dated failure modes and hook-enforces the fix).
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 5/5
- Output vs. baseline 9/10
- Docs & honesty 5/5
What Lore does
Runs a deterministic Node engine over your .specstory/history transcripts (any agent) to surface reproducible workflows with recurrence counts, next-reply outcome labels, and path:line evidence, then walks you through forging the good ones into installed SKILL.md packages. Triggers on "what could I make into a skill", "mine my lore", "forge skills from my history", or pointing at a .specstory/history directory.
How to install Lore
Claude Code plugin: /plugin marketplace add specstoryai/getspecstory, then /plugin install lore. Or copy the lore/ dir into your skills path. Needs Node >=22.5 (built-in SQLite); no external npm deps. Requires SpecStory to have recorded sessions to mine.
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Lore
-
/loreMines your SpecStory coding histories into corroborated, outcome-labeled skill candidates you forge into real skills.
It also activates on plain-language prompts like these:
-
What could I make into a skill from my history -
Mine my SpecStory history for reusable workflows -
Forge a skill from my past coding sessions
Frequently asked questions
- Is the Lore skill free?
- Yes. The skill itself is free from specstoryai/getspecstory. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Lore work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 20, 2026. Verdict: Tested · Works. Cloned the repo and actually ran the engine (pure Node stdlib, zero npm deps): index parsed 6 multi-agent sessions across 2 nested projects, and report surfaced the corroborated pattern "build:run × go build ▸ go test ▸ golangci-lint run" (3 sessions, 2✓/1✗ outcomes labeled from the user's next reply, exact commands from executed shell blocks, path:line refs, git author attribution) plus a verbatim pass-through summary — a clear, specific win over baseline transcript-guessing. Security scan clean: only local `git` via execFileSync for authorship, no network imports, no curl|sh, and the plan-validation hook correctly denies non-engine forge plans. All referenced scripts (mine-skills.mjs, hooks/validate-plan.mjs, workflows, lib/*) exist and `forged check` idempotency works. TRIGGER — SHOULD fire: "what could I make into a skill from my sessions?", "mine my lore and forge skills from my history", "turn my last 30 days of coding into reusable skills"; SHOULD NOT: "write me a skill for deploying to Vercel" (author-from-scratch, no mining), "summarize what I did in yesterday's session" (session summary, not skill extraction) — 5/5 correct. Only friction: requires the separate SpecStory recorder to have captured sessions, but the skill+engine themselves are fully self-contained and measured working. Exceptionally honest docs (names its own dated failure modes and hook-enforces the fix).
- What is the Lore SkillProof Score?
- 9.6/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 9/10, docs & honesty 5/5.
- How do I install Lore?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Lore with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.