# Protocol Source Assurance Matrix

**Last Updated:** 2026-04-07  
**Purpose:** Track whether the protocol contracts currently used by the repo are proven to be the approved upstream-native families from the proper official sources.

**Status meanings**

- `Assured`: repo wiring and verification clearly point to the approved upstream family
- `Partial`: some proof exists, but the current venue or contract family is still transitional, pilot-compatible, or incompletely verified
- `Not Assured`: the repo itself still treats the venue as provisional, pilot-only, or not yet reconciled to the upstream family

---

## 1. Matrix

| Protocol | Current repo posture | Assurance status | Notes |
|---------|-----------------------|------------------|-------|
| DODO V2 PMM | Canonical Chain 138 PMM stack, Mainnet public `cW*` PMM wave | Partial | Strongly wired in the repo, but Mainnet public pool replacements and full-surface migration are still in progress. |
| DODO v3 / D3MM | Live on Chain 138 as canonical pilot venue | Partial | On-chain health and planner visibility are good, but the repo still treats D3MM as pilot/private and the Blockscout verification check currently fails. |
| Uniswap | Chain 138 upstream-native lane plus broader pilot-compatible venue layer | Partial | The repo explicitly says upstream-native replacements remain a separate track for some surfaces. |
| Balancer | Pilot-compatible Chain 138 venue layer | Not Assured | Bytecode and liquidity are present, but the repo still labels this layer pilot-compatible rather than fully upstream-native / latest-family assured. |
| Curve | Pilot-compatible Chain 138 venue layer | Not Assured | Same posture as Balancer. |
| 1inch | Pilot-compatible Chain 138 router surface | Not Assured | Planner and venue references exist, but not as a fully proven latest official contract-family integration. |

---

## 2. Evidence from current repo state

### DODO V2 PMM

- Canonical stable PMM stack is still the default public PMM layer on Chain 138.
- Mainnet public `cW*` PMM rows are live, but the repo now carries migration work specifically because some pools expose partial DODO surfaces rather than full standard DODO behavior.

References:

- [PMM_DEX_ROUTING_STATUS](PMM_DEX_ROUTING_STATUS.md)
- [MAINNET_STANDARD_DODO_POOL_MIGRATION_PLAN](../03-deployment/MAINNET_STANDARD_DODO_POOL_MIGRATION_PLAN.md)

### DODO v3 / D3MM

Verified on 2026-04-07:

- `bash scripts/verify/check-dodo-v3-chain138.sh`
  - passed version / vault / oracle / quote checks
  - failed on Blockscout source-verification assertion for `D3Oracle`
- `bash scripts/verify/check-dodo-v3-planner-visibility-chain138.sh`
  - passed
  - `provider=dodo_v3`
  - `live=true`
  - `quoteLive=true`
  - `executionLive=true`

Conclusion: D3MM is operationally real, but source-assurance remains incomplete.

### Uniswap / Balancer / Curve / 1inch

Verified on 2026-04-07:

- `bash scripts/verify/check-chain138-pilot-dex-venues.sh`
  - bytecode present for all configured venues
  - liquidity present for all configured venues
  - failed on the planner-capability parsing check even though the dedicated D3 planner visibility verifier passed separately

The repo’s own docs still describe this layer as:

- live on Chain 138
- funded / planner-visible
- **pilot-compatible**
- with upstream-native replacements still optional future work for some venues

Reference:

- [DEX_AND_CROSS_CHAIN_CONTRACTS_NEEDED](DEX_AND_CROSS_CHAIN_CONTRACTS_NEEDED.md)

---

## 3. Required remediation

### DODO V2 PMM

- Replace or migrate partial-surface Mainnet pools to full-surface canonical DODO pools.
- Reconcile integration mappings, env references, docs, and deployment graph rows to the replacement addresses.

### DODO v3 / D3MM

- Fix Blockscout/source-verification posture for canonical D3 contracts.
- Promote D3MM from pilot/private to canonical public execution where policy allows.
- Extend beyond the current `WETH10 <-> USDT` pilot lane.

### Uniswap / Balancer / Curve / 1inch

- Build a protocol-by-protocol upstream assurance table with:
  - official upstream docs / repo
  - approved contract family
  - exact address in use
  - proof or gap
- Replace pilot-compatible venue contracts where policy requires latest upstream-native families.
- Reconcile planner-capability verification so the venue verifier and planner visibility tests tell the same story.

---

## 4. Current tested blockers

- D3MM health is good, but Blockscout/source assurance is not yet clean.
- The venue verifier succeeds on bytecode and liquidity but currently fails on the planner-capability response parsing step.
- Mainnet DODO public pools still require migration / audit work before they can be treated as fully assured standard DODO venues everywhere.