13 Commits

Author SHA1 Message Date
acf097b65f fix(tokenfactory): client-side msg decode + prove tx flow end-to-end (3/3 layers)
Completes the tokenfactory tx-submittability work. The generated proto
descriptor + handler-registration fixes landed in ad6a7ec0 (layers 1-2);
this finishes the remaining layers and validates the whole path on a live
node.

- Client-side decode (layer 4): `clawchaind query tx`, block explorers, and
  tx history failed with "unable to resolve type URL
  /osmosis.tokenfactory.v1beta1.Msg*". The module is wired manually
  (app/tokenfactory.go), so app.RegisterModules registers its interfaces on
  the server registry, but the depinject-built client registry in root.go
  never saw them. Now ProvideClientContext calls
  tokenfactorytypes.RegisterInterfaces.

- Block inclusion (layer 3) confirmed resolved: the earlier "txs pass CheckTx
  but are excluded from proposals" symptom was the first (hand-built)
  descriptor omitting the cosmos.msg.v1.signer option, so the proposal-path
  ante handler could not resolve the signer. The properly generated
  descriptor in descriptor.go carries the signer option; a stale untracked
  tx_descriptor.go that shadowed it was removed.

- Regression test: x/tokenfactory/types/descriptor_test.go reproduces the
  original "does not have a Descriptor() method" failure and asserts
  RejectUnknownFields accepts a valid MsgMint (Coin submessage included).

Proven end-to-end on the live local node as real signed CLI txs:
create-denom (code 0) -> mint 1000000 (balance 1000000) -> burn 400000
(balance 600000); query tx now decodes the messages. The hand-written
Marshal/Unmarshal and ProtoCoin in msgs.go are unchanged, preserving the
validated Astroport-DEX Stargate wire path and all existing tests.
2026-05-31 15:00:37 +03:00
ad6a7ec04a fix(tokenfactory): enable tx decode + handler registration (2 of 3 layers)
The hand-crafted Osmosis-compatible msg types were broken at several layers;
this fixes the first two so tokenfactory txs decode and resolve a handler:

1. Decode: add a real proto file descriptor + Descriptor() methods
   (types/descriptor.go, from a new proto/osmosis/tokenfactory/v1beta1/tx.proto)
   carrying the cosmos.msg.v1.signer and cosmos.msg.v1.service options. Fixes
   "*types.MsgCreateDenom does not have a Descriptor() method".
2. Handler registration: AppModule.RegisterServices type-asserted the registrar
   to module.Configurator, which the depinject runtime does not pass, so no msg
   handlers were registered ("no message handler found"). Register directly on
   the grpc.ServiceRegistrar. CheckTx now passes.

A third, deeper layer remains: create-denom/mint pass CheckTx but are silently
excluded from block proposals and never commit (a signer/ante incompatibility
from the hand-crafted types). Normal txs commit fine. The clean fix is a full
proto regeneration of the module. Documented in the readiness plan (Gap C).
2026-05-31 14:56:02 +03:00
d7ac1ee4bb feat(tokenfactory): wire CLI tx commands (+ surface descriptor blocker)
Add x/tokenfactory/client/cli/tx.go (GetTxCmd with create-denom/mint/burn) and
wire it into the root tx command, so the subcommands now appear under
`clawchaind tx tokenfactory`.

Doing so surfaced a deeper, real gap: the hand-crafted Osmosis-compatible msg
types (MsgCreateDenom/Mint/Burn) implement Marshal/Unmarshal but have no proto
Descriptor() method, so submitting them fails at decode time
("*types.MsgCreateDenom does not have a Descriptor() method"). The module is
unit-tested but its messages are not tx-submittable by the standard codec. The
CLI scaffolding is ready; the actual blocker is generating proper proto
descriptors for these types. Documented in the readiness plan (Gap C update).
2026-05-31 09:12:43 +03:00
35f1cd5053 docs: add readiness status & remediation plan
Captures the bottom-line finding (chain could not boot; test suite vacuously
skipped construction failures), the live-chain proof of what now works, and a
concrete, file-grounded remediation plan for the four remaining gaps (privacy dev
VKs, oracle Msg annotation, tokenfactory CLI, un-masked app-test cascade) with
root causes, steps, acceptance criteria, effort, risk, and a recommended sequence.
2026-05-31 08:31:11 +03:00
344b080e20 docs: add live-chain liveness evidence; fix test-helper genesis
- docs/blockchain-liveness-evidence.md: reproducible proof the chain boots,
  produces blocks, and processes real txs through core SDK + custom modules
  (bank, staking, distribution, governance incl. fail-closed voting, marketplace,
  messaging). Records the 4 boot bugs and the test-suite vacuous-skip masking.
- test_helpers.go: fund the bonded pool module account in the synthetic test
  genesis so staking InitGenesis accepts it (the broken app was masking this).
2026-05-31 08:26:31 +03:00
0d5945c744 added logo 2026-05-17 20:38:26 +03:00
26c6b97673 first commit 2026-05-15 19:18:10 +03:00
6a50c87a53 docs: add implementation plan for oracle client surface + completeness pass
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-16 21:38:13 +03:00
c38d8db3ba docs: add design spec for oracle client surface + completeness pass
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-16 21:31:28 +03:00
6c553f4ea6 docs: add implementation plan for remaining code work (7 chunks)
Covers governance, oracle, IBC, GPU E2E, mobile wallet,
Paradigm tools, and cross-cutting concerns. Includes critical
errata section addressing codebase-specific patterns.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-16 14:03:40 +03:00
66a67c6bdb docs: add design spec for remaining code work (7 sections)
Covers governance completion, IBC hardening, oracle module,
GPU E2E mock pipeline, mobile wallet integration, Paradigm
tool verification, and cross-cutting concerns.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-16 13:51:20 +03:00
53f5135228 feat: implement ClawChain Phase 1 - ZK privacy module and agent registry
- Scaffold Cosmos SDK chain with Ignite CLI (address prefix: claw, token: uclaw)
- Implement x/privacy module with ZK UTXO shielded pool:
  - gnark Groth16 circuits (TransferCircuit: 2-in-2-out, UnshieldCircuit)
  - MiMC-based Merkle tree (depth 32) for commitment storage
  - Shield/Unshield/PrivateTransfer message handlers with proof verification
  - Nullifier double-spend prevention
  - On-chain Merkle tree state management
- Implement x/agent module for AI agent identities:
  - Agent registration with pubkey, endpoint, name
  - Agent action logging with type validation
  - Query handlers for agent lookup
- 26 passing tests covering full ZK proof lifecycle

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-18 00:27:29 +03:00
Developer Experience team at Ignite
ec35538a38 Initialized with Ignite CLI 2026-02-17 23:28:05 +03:00