co-arising · shared knowledge, accountable change

Make domain knowledge shared,
implementations verifiable,
changes traceable.

How do we keep a business decision connected to the code that implements it? Write the concepts, rules and behavior in a readable truth source. Generate the supported implementation, keep integration explicit, and trace changes back to their declarations.

The Co-arising tree, rooted in human and business needs, branching into domain concepts and implementation.
Co-arising · The root tree of domain truth

Co-arising community

Grow shared understanding, together.

Discuss domain modeling, truth sources and event-sourced systems. Bring a question, share an experiment, or help shape the tools through practical feedback.

Join the Discord community
01

Keep the decision and its implementation connected

The smallest useful loop is truth tree → compiled products → generated code → source anchors. Each part answers a different question; no single green gate proves the business decision right.

Describe

What must remain true?

Use a file-based truth tree to name aggregates, commands, events and invariants. People can review the business rules before discussing implementation details.

Explore

What depends on this?

Query declared relationships in a local SQLite knowledge graph. Explore the bank snapshot; the graph can only show dependencies that have been modeled.

Generate

Which behavior is declared?

Supported guards, event bindings and state folds become code alongside CQRS/ES structure. Generation does not invent missing business policy.

Integrate

What remains handwritten?

Technical bindings and a separate handwritten zone connect the generated project to external effects and host services. Their responsibilities and tests stay explicit.

Trace

Where did this implementation come from?

Source anchors connect knowledge to generated code at the precision supplied. Verification checks freshness and bytes; it is not proof that every line is correct.

Review

Who approves a change?

People and agents can query, propose and test. Humans approve the intended behavior. Technical passes, independent experience and delivery remain separate on the status ledger.

02

Make one change in your own project

Use Node ≥ 22.13 in a new exercise directory. Acquiring the source checkout needs a network; the Counter generation loop needs no JVM or database service. The full walkthrough — including the bank example and the MCP integration — is on the hands-on guide.

Expand the first-generation commands
export TRUTHSRC_WORKSPACE="$PWD"
git clone https://github.com/co-arising/metarch-truthsrc.git
export TRUTHSRC_CLI="$TRUTHSRC_WORKSPACE/metarch-truthsrc/bin/truthsrc.js"
node "$TRUTHSRC_CLI" init counter-lab --format human
cd counter-lab
node "$TRUTHSRC_CLI" generate-project --project truthsrc.project.json --out generated --plan --format human
node "$TRUTHSRC_CLI" generate-project --project truthsrc.project.json --out generated --format human
node "$TRUTHSRC_CLI" query-project --project truthsrc.project.json --out generated --knowledge SetCount --format human
node "$TRUTHSRC_CLI" verify-project --project truthsrc.project.json --out generated --format human
03

Choose the question you came with

Insights

Chinese-language editorials on AI coding and co-arising’s engineering direction. These are perspectives, not verification of shipped capabilities; dated evidence stays on the status ledger.

Industry perspective

Architecture governance as AI coding evolves

A look at the shift from vibe coding toward agentic engineering: specifications, feedback loops and deterministic guardrails.

Engineering direction

What the shift could mean for co-arising

Exploring compact domain declarations, bounded agent work and model–code traceability as engineering directions, not a delivered-workflow claim.

Browse insights · articles in Chinese

04

Evidence, with its limits in view

The tree, code and anchor counts below retain their 2026-11-11 capture in SITE-FACTS; they are not new service-run counts. The engines are now published on Maven Central F-23. npm publication, hosting and BDD derivatives remain on the ledger.

157
truth atoms in the bank worked example F-02
149
named invariant declarations in the bank source F-09
235
generated files; build: 0 compile errors F-04 F-06
792
verified source anchors F-05
BANK TECHNICAL RESULT

On 2026-11-12, bank030-relocated-20261112-02 passed and rechecked source / anchors / code / service F-24, including real database/HTTP and automated browser observations F-26. This is candidate-specific technical acceptance, not independent experience (DEF-07), public matched-material delivery (DEF-15) or human usability acceptance (DEF-16). The old service02 failure remains historical (DEF-06). Read the acceptance scope and prerequisites.

PHILOSOPHY

Make the evidence boundary visible. These pages have no build step or npm dependencies; claims link to dated evidence and open items. A summary of privately reviewed reports is not a public reproduction package. Read why.

Next: follow a business question. See why payment and investment success are different facts.