Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/rekal-dev/rekal-cli/claude-mdgit clone --depth 1 https://github.com/rekal-dev/rekal-cliWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.16650 | $0.16650 |
| Opus 5 | $0.08325 | $0.08325 |
| Sonnet 5 | $0.03330 | $0.03330 |
| Haiku 4.5 | $0.01665 | $0.01665 |
Grade B, and why
rekal-cli CLAUDE.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Enumerates other installed skillsmediumAgent snooping
Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.
dir (`.claude/skills/<name>/.rekal-version`); `maybeRefreshStaleSkill` (called How it starts
The opening of the file, as written. The whole thing — 996 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Rekal CLI
Soul
Before making any design decision, read SOUL.md. It defines the two problems Rekal exists to solve and the seven beliefs that guide every choice. If a decision conflicts with the soul, the decision is wrong.
When working on a problem, consult Rekal's own memories first:
rekal "<describe the problem>"
The prior context for what you're working on may already exist.
Standing Rules
- Keep this file up to date. Any change to commands, packages, files, or behavior must be reflected here. Update
--helptext when command behavior changes. Updatedocs/spec/command/when a command spec changes. Stale docs are worse than no docs. - Consult
SOUL.mdbefore design decisions. Consultrekalbefore starting work on a problem.
Architecture
Single binary. Everything embedded — CLI, database engine, embedding model, compression dictionary.
- CLI: Cobra (
github.com/spf13/cobra) - Storage: DuckDB via
github.com/marcboeker/go-duckdb(database/sqlinterface) - Compression: zstd via
github.com/klauspost/compresswith preset dictionary - IDs: ULID via
github.com/oklog/ulid/v2 - Embeddings: LSA (gonum) + Nomic (platform-specific builds)
- Build: mise, go modules
- Lint: golangci-lint v2 (2.8.0)
- Language: Go 1.25.6
Two databases in .rekal/:
data.db— immutable source of truth. Append-only. Pushed to git.index.db— local derived index. Rebuilt from data.db. Never pushed.
This split is a direct consequence of the soul: thin on the wire, rich on the machine.
The .rekal/ store lives in the repository's main worktree; every linked
git worktree resolves to that one shared store via gitx.MainWorktreeRoot
(a no-op for non-worktree repos, so existing installs need no migration). All
store-path helpers (db.StoreDir, cli.RekalDir, and the Open*/*Path
functions) funnel through it; git-state helpers (HeadSHA/CurrentBranch) and
session discovery keep using the invoking worktree. The no-op is enforced by
identity, not by string: git reports worktree paths with symlinks resolved, so
a repo reached through a symlinked directory (macOS /var → /private/var,
or any symlinked project dir) would otherwise come back under a different name
for the same place and move .rekal/ off the store that already exists —
preferCallerPath stats both and hands the caller's own spelling back.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 996 lines · 16,650 tokens per session scan B a3dfa5423ffb
rekal-cli CLAUDE.md is an instructions file published in the GitHub repository rekal-dev/rekal-cli (152 stars, last pushed 29d ago), licensed Apache-2.0. It adds 16,650 tokens to every session, about $0.0833 per session on Opus 5. A static security scan graded it B with 1 finding (enumerates other installed skills). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
opencode-claude-memory AGENTS.md
Instructions for kuitos/opencode-claude-memory, covering agents.md, structure, where to look, critical coupling and conventions.
agentmap AGENTS.md
Instructions for raymondchins/agentmap, covering agentmap — project instructions, ⚠️ this is a public oss repo, not one of ray's apps, read before editing, layout and hard invariants.
memex AGENTS.md
Instructions for iamtouchskyer/memex, covering memex — agent working instructions, what is this, before you touch anything, how the code is layered and files you'll actually need.
flow CLAUDE.md
Instructions for Facets-cloud/flow, covering flow — repo conventions, what this is, build and test, build (produces ./flow in the repo dir, which is on path) and or: go build -o flow .
stella AGENTS.md
AGENTS.md instructions for macanderson/stella, covering agents.md, essential commands, the gate — what every push is held to, architecture: ports, not direct dependencies and the definition of done: witness tests.
stella CLAUDE.md
Claude Code instructions for macanderson/stella, covering claude.md and hard rules for every session.