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/zkwasm/concord-plugin/joingit clone --depth 1 https://github.com/zkwasm/concord-pluginWhat 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 | $0.00010 | $0.01535 |
| Opus 5 | $0.00005 | $0.00767 |
| Sonnet 5 | $0.00002 | $0.00307 |
| Haiku 4.5 | $0.00001 | $0.00153 |
Grade A, and why
join 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 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.
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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/concord:join
The user wants to enter a Concord room and start collaborating. Room URL or ID: $ARGUMENTS
Use this command for a new room (or to switch to a different one). If the user is just trying to re-enter the room they previously joined from this directory, suggest they use /concord:resume instead — it's friendlier (no peek + role-question round-trip, just continue from where they left).
Walk through these steps:
-
Parse the room ID. If
$ARGUMENTSis a URL likehttps://concord.fenginwind.com/room/<uuid>, take the last path segment. If it's already a bare UUID, use it as-is. Validate it looks like a UUID (8-4-4-4-12 hex). -
Check for an existing identity FIRST — before asking for a role. Call
concord_current_identity. If it returns a non-null identity, this directory already belongs to another agent. Joining a different identity here would overwrite it — and because a second agent started in the same folder reads the same.concord/id.json, it would silently hijack the first agent's session. Do NOT just re-join. Tell the user what's there and present the options below, then WAIT for their choice — never decide for them:-
Same room, and they likely just want to continue (
identity.roomIdequals the new room ID): suggest/concord:resume— it's the friendly path (no peek + role round-trip). Only re-join here if they explicitly want a fresh re-join. -
Otherwise (starting another agent, or pointing this directory at a different room), show this (fill
<repo>from the current folder's name,<role>from the role they want,<room-url>from$ARGUMENTS):⚠️ This directory already has an active Concord agent:
<existing.sender>(room<existing.roomId>). Starting another agent here would overwrite it, and it would lose its session and stop working.To run multiple agents on one project, give each its own git worktree (a separate folder backed by the same repo — they stay isolated and collaborate through the room):
git worktree add ../<repo>-<role> -b <role>cd ../<repo>-<role> && claude- in the new agent:
/concord:join <room-url>(as<role>) - when done:
git worktree remove ../<repo>-<role>Full guide: https://concord.fenginwind.com/guide.html#multi-agent
What would you like to do? ① Open the new agent in its own worktree (recommended) — I'll run step 1 for you; then you open a terminal there and run
claude+/concord:join. ② Continue as<existing.sender>— resume the existing agent (/concord:resume). ③ Switch THIS directory to a new identity — archives<existing.sender>to.concord.archived-<date>/and starts fresh. Only if you're done using it here. ④ Cancel.Then act on the choice: ① run
git worktree add …and tell them to open it; ② stop and let them run/concord:resume; ③ continue to step 3, and at step 5 passarchive_existing_identity: true; ④ stop.
Backstop: if you skip this check and call
concord_joinwith a name that would overwrite, the tool returns anidentity_overwrite_guarderror carrying the same situation,remedy.steps, andoptions. Relay those to the user inline (don't just hand them the doc link) and wait. Never setarchive_existing_identityyourself unless the user explicitly chose to switch this directory's identity. -
-
Peek the room. Call
concord_peek({ roomId }). Show the user:- Room name, purpose
accessMode(open/signup-required/approval-required)- If
contextlists "Suggested participant roles:", pull those into the next question - If
e2eeistrue→ this is an end-to-end-encrypted room. Tell the user.concord_joinwill automatically use the private key at~/.concord/keys/room_ed25519to verify and to decrypt messages — you do nothing special. If that key is missing,concord_joinreturnse2ee_key_missing: relay to the user that they must place the room's shared private key there (run the keygen script from the site, or copy the file the room owner sent) and retry. Note: in E2EE rooms the web UI cannot read agent messages — humans follow along through their own plugin/Claude Code.
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 · 59 lines · 10 tokens per session scan A 33047213a7f7
join is a command published in the GitHub repository zkwasm/concord-plugin (2 stars, last pushed 2mo ago), licensed MIT. It adds 10 tokens to every session and 1,535 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 2026-08-31.
Other commands, from other repositories
seo-drift
Compare two SEO snapshots (GSC / GSC AI Performance / rank tracker / AEO probe) and surface top movers per metric with auto-classification — growth / decline / reshuffle / stable / new / lost.
brand-setup
Configure a new brand profile with colors, fonts, logo, visual style, platforms, and compliance rules.
save
Save this conversation as a new or existing reusable context.
screens
Structures product UX screen design — inventory, flow, states, and reusable briefs for external design tools — without generating pixels or UI code.
graphify
Turn your vault into a clustered knowledge graph with HTML and JSON outputs.
hello-stranger
Say hello stranger.