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/justinjdev/fellowshipWrote 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/justinjdev/fellowship/rekindle)<a href="https://agentmods.dev/commands/justinjdev/fellowship/rekindle"><img src="https://agentmods.dev/badge/commands/justinjdev/fellowship/rekindle/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/justinjdev/fellowship/rekindle"><img src="https://agentmods.dev/badge/commands/justinjdev/fellowship/rekindle.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00043 | $0.01497 |
| Opus 5.5 | $0.00017 | $0.00599 |
| Sonnet 5.5 | $0.00009 | $0.00299 |
| Haiku 4.5 | $0.00004 | $0.00150 |
Grade A, and why
rekindle 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 1mo 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 — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Rekindle — Fellowship Crash Recovery
Overview
Reconstructs fellowship state from on-disk artifacts after a session crash and transitions into Gandalf coordinator mode with recovered context. The flame that was quenched can be rekindled.
When to Use
- Session crashed or context window filled up during a fellowship
- User returns to find scattered worktrees from a previous fellowship
- User invokes
/rekindledirectly
Process
Note:
.fellowship/is the default data directory. Users can override it viadataDirin~/.claude/fellowship.json. AllfellowshipCLI commands and paths below use the configured data directory automatically.
Step 1: Scan
Run the CLI to discover fellowship artifacts:
~/.claude/fellowship/bin/fellowship status --json
This reads the fellowship and quest state recorded in the fellowship database, checks each worktree for a checkpoint (.fellowship/checkpoint.md), and detects branches already merged into the fellowship's base branch.
If no quests are found, report: "There is nothing to rekindle. The ashes have gone cold." and stop.
Step 2: Classify
Each quest gets one classification:
| Classification | Condition | Action |
|---|---|---|
| Shipped | Branch merged into the base branch | Skip — already shipped |
| Resumable | Has .fellowship/checkpoint.md |
Continue from current phase with checkpoint context |
| Stale | No checkpoint | Restart current phase from scratch |
Step 3: Present Recovery Dashboard
Show the user what was found:
The flame that was quenched can be rekindled.
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
quest-api-auth │ Implement (checkpoint ✓) │ Resumable
quest-db-schema │ Plan (checkpoint ✓) │ Resumable
quest-ui-login │ Research (no checkpoint) │ Restart phase
Merged (skipping):
fellowship/config-fix (branch merged into main)
Proceed with recovery? (y/n)
If the user declines, stop. Do not proceed without confirmation.
Step 4: Re-spawn Fellowship
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.
- 1mo ago Changed · -1 lines c3e72f1c38c0
- 1mo ago Changed · -64 lines f3e109dede7f
- 1mo ago First seen · 174 lines · 43 tokens per session scan A 19ebb9098f59
rekindle is a command published in the GitHub repository justinjdev/fellowship (5 stars, last pushed 15d ago), licensed Apache-2.0. It adds 43 tokens to every session and 1,497 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-31.
Other commands, from other repositories
session-validate
End-of-session self-training — captures session metrics, extracts patterns via metaclaw-learner, updates instincts, and generates optimization hypotheses for the next run.
context-engineer
Context engineering agent — researches context window optimization from arxiv, builds token-efficient context packages for downstream agents, manages cross-session persistence via MetaClaw.
pm-crystallize
Promote a project-local lesson to /.great-pm/decisions.md (cross-project memory) when it has 3+ hits with high confidence. Surfaces candidates as PROPOSALS — never auto-promotes; human approval required.
memories
View and manage learned memories.
forget
Delete specific memories.
learn
Add new learning to memory.