Author: Zijun Fu | Produced by the Civilization Leap Research Group
Let Seeing Change the Next Step
Start from a specific question, find a usable tool, and bring back the difficulties you encounter.
Continued existence is the baseline. Tools help us see consequences before acting and retain opportunities to correct decisions afterward. Their value must ultimately be judged by whether people's situations change.
This page provides navigation and feedback drafts. Public submissions go through project GitHub pages; private correspondence goes through email. Filling in this page does not automatically send or upload anything to this website.
What do you want to understand now?
How can we see overlooked people and consequences before acting?
Human-COS
Reveal possible results, causal paths, others' situations, and remaining room for correction earlier, so that new understanding can enter purpose formation.
The public Core provides S0/S1 foundations. FFT-1 is a fixed synthetic Mock trial. The current public launcher does not analyze your real-world case.
Read the English operator guide first and confirm that this is a fixed Mock.
With Docker available, use the repository's launcher; without it, choose read-only review.
Record the version, steps actually executed, and findings. Do not record read-only review as execution.
A successful run supports only the tested process; it does not prove that real-world judgment is effective.
Who benefits from an arrangement, and who bears its costs?
CCM — Computable Cooperation Mechanisms
Bring multiple subjects, third parties, non-tradable boundaries, exit conditions, and unknowns into one discussion, so arrangements can be compared, challenged, and renegotiated.
Public reference implementation v0.1.1; recalculates declared inputs and constraints.
Start with the five-minute threshold example to see how rules change the determination.
Use the three-minute problem card to identify parties, costs, non-negotiable conditions, and unknowns.
Run only what can be represented without distortion. Report what cannot be represented as a model gap.
It does not decide purposes for the parties, compress incomparable losses into an aggregate score, or prove that inputs are true or rules legitimate.
Can a subject still correct, refuse, and recover choice?
CI-001 — Irreversible Closure Criterion
Before structural channels are locked, examine whether a specific subject's capability still has executable paths for correction, exit, and recovery within the relevant window.
Stable version V1.4; V1.5-RC2 is a public review candidate. Use the two version paths separately.
Select the stable version or review candidate in the repository and start with a completed example.
Specify one subject, one capability, one scope, and one window, then record C/E/R.
Retain ? when evidence is insufficient. If all three currently fail, separately examine reconstructability and provide clauses and evidence with feedback.
The form grants no enforcement authority. Unknown does not mean failed; self-assessment does not mean independent certification.
Functional descriptions follow repository materials made public or accessed with authorization on October 4, 2026. After repository updates, follow the documentation for the version you use. This page checked entry points; it did not rerun the four tools.
Turn one use into the next improvement
Explaining which step failed to address your problem helps more than simply saying good or bad. Report failures, ambiguities, alternatives, or how the tool changed your judgment.
After submitting, how can you follow the response?
After confirming submission on GitHub, save the Issue number and URL generated by the platform. Responses, supplementary evidence, and closure records remain there. Merely generating a draft here or opening GitHub does not count as submission.
Email must be sent from your email application. This page cannot confirm delivery or generate a receipt number. Keep your sent message and follow up in the original thread. Immediate replies are not promised at this stage.
Maintainers' responses should state how they understand the issue, whether evidence is sufficient, what they plan to change or why they will not yet change it, and link actual revisions. Closing an issue does not automatically prove resolution; new evidence can still be raised.
If using a tool merely adds a report while nobody can change the next step, the tool still has unfinished work. It should help people retain a future, not demand their obedience.
This page translates the complete existing public HTML text. Linked works and files retain their own publication and version boundaries; unavailable English materials are marked explicitly.