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/terrene-foundation/kailash-coc-claude-pyWrote 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/terrene-foundation/kailash-coc-claude-py/ecosystem-init)<a href="https://agentmods.dev/commands/terrene-foundation/kailash-coc-claude-py/ecosystem-init"><img src="https://agentmods.dev/badge/commands/terrene-foundation/kailash-coc-claude-py/ecosystem-init/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/terrene-foundation/kailash-coc-claude-py/ecosystem-init"><img src="https://agentmods.dev/badge/commands/terrene-foundation/kailash-coc-claude-py/ecosystem-init.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 | $0.00041 | $0.02294 |
| Opus 5 | $0.00020 | $0.01147 |
| Sonnet 5 | $0.00008 | $0.00459 |
| Haiku 4.5 | $0.00004 | $0.00229 |
Grade A, and why
ecosystem-init 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.
How it starts
The opening of the file, as written. The whole thing — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/ecosystem-init — Onboard an Ecosystem (once per fork)
The once-per-fork ceremony that configures a NEW ecosystem (canon, or a client fork that copied the
loom↔build↔use ecosystem). One of the three onboarding surfaces — distinct from /enroll (once per
operator) and /onboard (read-only, every session); see (loom-internal reference).
The three share ZERO write-authority: /onboard writes nothing, /enroll writes roster + local-links,
/ecosystem-init writes the ecosystem-config + genesis trust-root.
Usage: /ecosystem-init (no args; runs in the current fork's loom checkout)
Strictly dependency-ordered BEFORE this command, for a client fork: /clean-instantiate — the
once-per-clone CLEAR ceremony (.claude/commands/clean-instantiate.md) — MUST have run to a 0-exit
fail-closed assert-zero gate on any repo templated or cloned FROM canon. C3 below re-anchors genesis to
the CLIENT's owner; it MUST NOT run over a clone that still carries canon's trust-identity (roster,
journal, ecosystem.json) at rest — re-anchoring on top of un-cleared identity leaves canon's substrate
readable underneath the new root. Canon itself (never cloned FROM anything) has no such precursor and
runs /ecosystem-init directly. For a client whose clone is not yet clear, /ecosystem-init is NOT the
first command to run — /clean-instantiate is.
Strictly dependency-ordered after this command: an operator runs /enroll (identity), then /onboard
(reads each session). /ecosystem-init does NOT silently enroll the initiating operator — it hands off
to /enroll (C5). Procedure detail (input prompts, the D6 schema field set, the disclosure-scan
invocation shape, the genesis-ceremony call) lives in .claude/skills/43-ecosystem-init/SKILL.md per
cc-artifacts.md Rule 3; this command is the entry point.
The five invariants (load-bearing — the redteam surface)
- C1 disclosure gate fires BEFORE the ecosystem-config write. The config names real org slugs
(the #255/#252 disclosure class).
node .claude/bin/scan-synced-disclosure.mjs --check --root <fork-checkout>MUST run and exit 0 BEFORE.claude/bin/ecosystem.jsonis written; ANY finding → HALT, genericize + relocate, re-scan. Placement does not proceed on a non-zero exit (theartifact-flow.mdIntake-Scrub shape, applied at ecosystem-config write time). The scan covers the SURROUNDING synced surface, NOTecosystem.json's own slugs (the scan runs BEFORE the config is written, and the scanner self-excludes the config at loom-source) — those slugs are fenced by PATH (invariant 5: loom-only never-synced + publish-excluded), not by this scrub (the SKILL § "Disclosure isolation" carries the full nuance). - Human gate confirms the org slugs. For a CLIENT fork, the human MUST confirm each slug points at
the CLIENT's org, never canon's — a fork AUTHORS its OWN
ecosystem.json(it never inherits canon's; the structural basis of cross-ecosystem isolation, D6 plan §A1). Automated org-slug placement is BLOCKED. - Genesis trust-root is established via the EXISTING ceremony. C3 invokes
.claude/hooks/lib/genesis-ceremony.js::runEnrollmentCeremony(the org-owned-bootstrap path,multi-operator-coordination.md§6 + §1 issue-#358 relaxation) — it does NOT re-implement genesis. Owner-class gate; fail-CLOSED (any failed gh-api verification refuses to anchor). On a client fork, C3 MUST NOT run until/clean-instantiate's fail-closed assert-zero gate has exited 0 (see above) — the CLEAR precedes the establish-new-root step it gates. - C5 does NOT silently enroll the initiating operator. The ceremony ends by handing off to
/enroll; enrolling the operator is/enroll's job, gated separately (knowledge-convergence.mdMUST-5 forbids/onboard-class commands auto-running roster/genesis writes for an operator). ecosystem.jsonis ecosystem-private by path. It is written to.claude/bin/ecosystem.json— committed-but-never-synced (sync-manifest.yaml::loom_only), never-published (publish-to-public.mjsEXCLUDE+KILL), and scanner-self-excluded ONLY at loom-source (REPO_ROOT_ACTIVE === REPO_ROOT); a DESTINATION--rootscan of another repo DOES scan a stray copy, failing loud on bare org slugs — the belt-and-suspenders backstop. A fork carries its OWN file; no canon→client sync path exists, so canon org slugs cannot travel into a client (D6 plan §4 fence-i).
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.
- 6d ago First seen · 127 lines · 41 tokens per session scan A a4967036d4e3
ecosystem-init is a command published in the GitHub repository terrene-foundation/kailash-coc-claude-py (12 stars, last pushed 22d ago), licensed Apache-2.0. It adds 41 tokens to every session and 2,294 once invoked, about $0.0002 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 2026-09-03.
Other commands, from other repositories
template
Manage issue templates for streamlined issue creation.
sync-linear
Sync current work with Linear ticket status.
add-note
Add an internal or external note to a ConnectWise PSA ticket.
fest-show
Show festival progression (in-progress tasks, roadmap, and dependency view).
dispatcher
Pick the next-best repo to work on across the portfolio — rank free repos, recommend one, claim its lease atomically, and route to the entry command.
workpm
A project-management workflow for coordinating multiple AI workers through five stages. It includes task assignment, shared activity logs, worker replacement, and final checks.