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 commands/flinedev/recall/sessiongit clone --depth 1 https://github.com/FlineDev/RecallWhat 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 | $0.00004 | $0.02135 |
| Opus 5 | $0.00002 | $0.01068 |
| Sonnet 5 | $0.00001 | $0.00427 |
| Haiku 4.5 | $0.00000 | $0.00214 |
Grade B, and why
session scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
grep -rl "\"customTitle\":\"$ARGUMENTS\"" ~/.claude/projects/*/?.jsonl ~/.claude/projects/*/??.jsonl ~/.claude/projects/*/*.jsonl 2>/dev/null | head -1 How it starts
The opening of the file, as written. The whole thing — 178 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Recall a Previous Session
If no session ID was provided (check $ARGUMENTS — if it's empty or missing), respond with:
Usage:
/recall:session <session-id or name>Pass either a UUID session ID or a custom session name (from
/rename). The session ID is printed when you exit Claude Code. You can also find recent sessions withclaude --sessions(run outside Claude Code).
Then stop — do not proceed with the steps below.
Recall a previous Claude Code session with much richer context than auto-compaction provides. Keeps ALL user messages and assistant responses verbatim while achieving 99%+ compression.
Background
Two hooks handle automatic recovery (no action needed):
- PreCompact parses the transcript and writes to
.claude/recall-context.md - SessionStart cleans up the recall file to prevent stale content
This command handles manual recall — when the user runs /recall:session <session-id> in a new session.
Execution Steps
Step 0: Resolve script paths
Resolve the Recall plugin's installed path from installed_plugins.json, then find the scripts:
RECALL_SCRIPTS="$(python3 -c "import json,pathlib; d=json.load(open(pathlib.Path.home()/'.claude/plugins/installed_plugins.json')); entries=d.get('plugins',{}).get('recall@FlineDev',[]); print(entries[0]['installPath']+'/skills/session/scripts' if entries else '')" 2>/dev/null)"
echo "$RECALL_SCRIPTS"
If empty, the plugin may not be installed. Tell the user to run /plugin install recall.
Step 0.5: Resolve session ID
The user provided: $ARGUMENTS
If it matches a UUID pattern (8-4-4-4-12 hex like a1b2c3d4-e5f6-7890-abcd-ef1234567890), use it directly as the session ID.
If it does NOT match a UUID (e.g. a custom name like my-session), it's a session name set via /rename. Resolve it to a UUID:
grep -rl "\"customTitle\":\"$ARGUMENTS\"" ~/.claude/projects/*/?.jsonl ~/.claude/projects/*/??.jsonl ~/.claude/projects/*/*.jsonl 2>/dev/null | head -1
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.
- yesterday First seen · 178 lines · 4 tokens per session scan B 8eb341c5b5f6
session is a command published in the GitHub repository FlineDev/Recall (10 stars, last pushed 5mo ago), licensed MIT. It adds 4 tokens to every session and 2,135 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.