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/assafkip/kipi-systemWrote 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/assafkip/kipi-system/design-chain)<a href="https://agentmods.dev/commands/assafkip/kipi-system/design-chain"><img src="https://agentmods.dev/badge/commands/assafkip/kipi-system/design-chain/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/assafkip/kipi-system/design-chain"><img src="https://agentmods.dev/badge/commands/assafkip/kipi-system/design-chain.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 | $0.00000 | $0.01407 |
| Opus 5.5 | $0.00000 | $0.00563 |
| Sonnet 5.5 | $0.00000 | $0.00281 |
| Haiku 4.5 | $0.00000 | $0.00141 |
Grade A, and why
design-chain scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 25d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Run one design round. Usage: /design-chain [project] [round].
project is a NAME, never a path: the project field of a design-chain.json under a root in
instance-registry.json (today: askconsulting). round defaults to today's date, with a letter
suffix if that folder exists (2026-09-19b). The founder does not remember round numbers.
python3 - "$ARGUMENTS" <<'EOF'
import json, os, pathlib, sys
want = (sys.argv[1].split() + [""])[0].lower()
reg = pathlib.Path(os.environ.get("CLAUDE_PROJECT_DIR", ".")) / "instance-registry.json"
roots = [pathlib.Path(i["path"]) for i in json.load(open(reg))["instances"]] if reg.is_file() else []
found = {}
for r in roots:
c = r / "design-chain.json"
if c.is_file():
try: found[json.load(open(c)).get("project", r.name).lower()] = r
except ValueError: pass
if want and want in found: print(found[want]); sys.exit()
if want: print("UNKNOWN project %r. Known: %s" % (want, ", ".join(sorted(found)) or "none")); sys.exit(2)
print("NO project given. Known: %s" % ", ".join(sorted(found)))
EOF
UNKNOWN or NO: stop and say so.
What holds each step
Every step below names the executable that checks it. design-chain-gate.py seal <round> runs the
producers itself and then the whole chain; nothing here is enforced by this document. Where a step
says a gate function, that function is in design-chain-gate.py and has its own test under
q-system/.q-system/scripts/test/.
Step 0: the vision comes before the brief
Say what the page is trying to BE, in the founder's own words, before any constraint. Held by
vision_problems(): with a vision block in the config, VISION.md exists, carries the founder's
quoted lines, brief.md quotes one of them verbatim, and a declared component has a spec marked
reviewed. Scar: three sealed directions, all wrong, every one assembled from constraints.
Step 1: read the owners, in full, this session
Read every owner file the config names, whole. Held by brief_read_problems(): when the config asks
for a fresh brief, the hook writes brief-reads.json from this session's transcript, and the seal
refuses a brief whose owners were not read in full, or a record copied from another round.
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.
- 25d ago First seen · 117 lines · 0 tokens per session scan A 75b2f8f656ad
design-chain is a command published in the GitHub repository assafkip/kipi-system (121 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,407 tokens. 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 2026-09-21.