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.
git clone --depth 1 https://github.com/roberto-mello/lavraWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/roberto-mello/lavra/lavra-checkpoint)<a href="https://agentmods.dev/commands/roberto-mello/lavra/lavra-checkpoint"><img src="https://agentmods.dev/badge/commands/roberto-mello/lavra/lavra-checkpoint.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00018 | $0.01149 |
| Opus 5 | $0.00009 | $0.00575 |
| Sonnet 5 | $0.00004 | $0.00230 |
| Haiku 4.5 | $0.00002 | $0.00115 |
Grade A, and why
lavra-checkpoint scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 176 lines — stays where its author put it; the contents beside it link to each section on GitHub.
<project_root>
All .lavra/ paths are relative to the project root. If you cd into a subdirectory during work, resolve the project root first:
PROJECT_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || echo "$PWD")
Then prefix all .lavra/ paths with "$PROJECT_ROOT/" when invoking them via Bash.
</project_root>
Step 1: Review Session Work
Scan the conversation history for:
- Code changes made:
git diffandgit diff --cached - Bugs fixed, features added, refactors performed
- Technical decisions made
- Insights discovered
Step 2: File Beads Issues
For each distinct unit of work done during this session that doesn't already have a bead:
bd create --title="<descriptive title>" --type=<task|bug|feature> --priority=2
If work is complete, close it immediately:
bd close <id>
If work is in progress, mark it:
bd update <id> --status=in_progress
Check existing issues first to avoid duplicates:
bd list --status=open
bd list --status=in_progress
Step 3: Add Knowledge Comments
For each filed or existing bead that was worked on, add at least one knowledge comment:
bd comments add <id> "LEARNED: <key technical insight>"
bd comments add <id> "DECISION: <what was chosen and why>"
bd comments add <id> "FACT: <constraint, gotcha, or environment detail>"
bd comments add <id> "PATTERN: <recurring convention or idiom>"
bd comments add <id> "INVESTIGATION: <root cause analysis>"
These will be auto-captured by the memory-capture hook.
Step 4: Commit Changes
If there are uncommitted changes:
- Check status:
git status - Stage only session-changed files (NOT
git add -A) - Commit with descriptive message
Step 5: Write Session State
Write .lavra/memory/session-state.md to preserve position awareness across context compaction:
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.
- 7d ago First seen · 176 lines · 18 tokens per session scan A f05b6b4cf114
lavra-checkpoint is a command published in the GitHub repository roberto-mello/lavra (50 stars, last pushed 2mo ago), licensed MIT. It adds 18 tokens to every session and 1,149 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
import
Import issues and memories from JSONL.
template
Manage issue templates for streamlined issue creation.
sync
Synchronize issues with a configured Dolt remote.
dep
Manage dependencies between issues.
decision
Record, list, and manage project decisions with rationale tracking.
plan-to-beads
Convert a Claude Code plan file into beads epic + tasks for cross-session tracking.