recall

recall is a skill for Claude Code, Codex from floomhq/moto. It costs 60 tokens per session (622 once invoked), scanned A, original, MIT.

A context-recovery guide for restoring an AI coding agent's work after its conversation has been shortened or compacted.

In plain words
What is it for?
Use it after compaction or when you need to reconstruct what the agent was doing from saved transcripts and work plans.
Why use it?
It helps recover the active task, completed steps, decisions, errors, and next actions without starting over.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: reads .claude/ paths; mentions CLAUDE.md.

Install

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.

agentmods
npx agentmods add skills/floomhq/moto/recall
Any agent
npx skills add floomhq/moto --skill recall
Clone the repo
git clone --depth 1 https://github.com/floomhq/moto

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for recall

README.md
[![agentmods](https://agentmods.dev/badge/skills/floomhq/moto/recall.svg)](https://agentmods.dev/skills/floomhq/moto/recall)
Your own site
<a href="https://agentmods.dev/skills/floomhq/moto/recall"><img src="https://agentmods.dev/badge/skills/floomhq/moto/recall.svg" alt="Measured on agentmods" height="20"></a>
Per session 60 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 622 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00060 $0.00622
Opus 5 $0.00030 $0.00311
Sonnet 5 $0.00012 $0.00124
Haiku 4.5 $0.00006 $0.00062

Measured 6d ago against content hash 1c412571081c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

recall 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 6d 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.

claude/skills/recall/SKILL.md · 98 lines

How it starts

The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Recall Skill

Recover working context after compaction. JSONL transcripts are NEVER deleted - they persist at ~/.claude/projects/*/. Compaction only clears the live context window.

Workflow

Step 1: Get Recent Messages

Use MCP tool (preferred):

recall_recent 5

CLI fallback:

session-recall --recent 10

Step 2: Find Active Task

Search for what was being worked on:

recall_search "workplan|task|fixing|implementing|debugging"

CLI fallback:

session-recall "workplan task fixing implementing"

Step 3: Find Active Workplan

find . ~/Downloads /root -name "WORKPLAN-*.md" -mtime -1 2>/dev/null | sort

Read the most recent one if found. It is the external brain - re-read it fully before continuing.

Step 4: Reconstruct Context

From the recalled messages and workplan, identify:

  • What task was in progress
  • What step was last completed
  • What the next step is
  • Any errors or blockers encountered

Use targeted searches if needed:

recall_search "error|failed|blocked|next step"
recall_decisions

CLI fallback:

session-recall --report
session-recall "error failed blocked"

Step 5: Persist Lessons (Optional)

If patterns or corrections were found in the session that should be saved:

recall_apply

MCP Tools Reference

Tool Purpose
recall_search "keyword" Search transcripts by keyword
recall_recent N Get last N human messages
recall_report Analyze errors, retries, corrections
recall_decisions Find key decisions made
recall_list List available sessions
recall_apply Persist lessons to CLAUDE.md/MEMORY.md

CLI Fallback Reference

session-recall "keyword"       # Search current session
session-recall --recent 10     # Last 10 human messages
session-recall --report        # Session analysis
session-recall --all 5         # Cross-session (last 5 sessions)

Key Fact

Transcripts are at ~/.claude/projects/*/. They are append-only JSONL files. Compaction does not touch them. Any content from any past session is recoverable. Never say "can't recover" compacted content.

Read the full file on GitHub · 98 lines

Changes

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.

  1. 6d ago First seen · 98 lines · 60 tokens per session scan A 1c412571081c

Subscribe to this mod's changes

recall is a skill published in the GitHub repository floomhq/moto (32 stars, last pushed 2mo ago), licensed MIT. It adds 60 tokens to every session and 622 once invoked, about $0.0003 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.