feat(contracts): governance benchmarks, insurance pool deploy profile, spec/ABI coverage

[brightpixel-dev]

Closes #522
Closes #523
Closes #524
Closes #525

What changed

#522 — Governance contract benchmark baselines

• Added contracts/iln_governance/src/tests_benchmarks.rs with execution-cost benchmarks for create_proposal, cast_vote, and delegate_votes, mirroring the existing invoice_liquidity/tests_benchmarks.rs pattern.
• Added contracts/iln_governance/benchmarks/baseline.json with measured baselines.
• Generalized scripts/check_benchmark_regression.sh to check a list of <contract-dir>:<baseline-file> pairs (now invoice_liquidity + iln_governance by default) instead of being hardcoded to one contract.
• Updated docs/benchmarks.md and CONTRIBUTING.md accordingly.

#523 — Docker Compose profile for local insurance pool testing

• scripts/deploy-local.sh was silently missing insurance_pool from its contract list — added it, so ./scripts/deploy-local.sh now deploys all 5 contracts as intended ("local dev needs all contracts available").
• Added scripts/deploy-insurance-pool.sh to deploy/initialize just the insurance pool.
• Added an opt-in insurance Docker Compose profile (insurance-pool-deploy service) to docker-compose.yml and docker-compose.test.yml — run with docker compose --profile insurance up insurance-pool-deploy. It's profile-gated (not part of the default docker compose up) since it installs the Stellar CLI via cargo install on first run, which can take several minutes.
• Added INSURANCE_POOL_CONTRACT_ID env var wiring to the indexer/notifications services, and documented all of the above in docs/local-development.md.

#524 — Contract spec JSON for insurance pool

• scripts/gen-spec.ts now also emits an insurance_pool spec, added under a new contracts key in docs/contract-spec.json.
• The existing top-level fields (which describe InvoiceLiquidityContract) are left untouched for backward compatibility with scripts/generate-sdk.ts, which reads them directly — verified via npx tsx --test scripts/generate-sdk.test.ts (5/5 passing).

#525 — ABI generation for insurance pool

• scripts/gen-abi.ts now loops over both contracts instead of being hardcoded to invoice_liquidity.
• While generalizing it, fixed two bugs that were producing an unusable docs/contract-abi.md (only 3 functions and 0 errors were ever captured for invoice_liquidity): the return-type regex broke on multi-word types like Result<(), ContractError>, and the errors regex only looked in lib.rs, but ContractError is defined in errors.rs. Both contracts now render their full function/error lists correctly.

Why

• Add benchmark baselines for governance contract #522 — no instruction-count baselines existed for the governance contract, so regressions there went unmeasured.
• Add docker-compose profile for local insurance pool testing #523 — the insurance pool contract had no way to be deployed/tested locally alongside the other contracts.
• Add contract spec JSON for insurance pool #524 — the insurance pool contract had no machine-readable spec for tooling/SDK codegen.
• Add ABI generation for insurance pool #525 — the insurance pool contract had no ABI documentation, and the existing generator for the other contract was also silently broken.

How to test

• cd contracts/iln_governance && cargo test benchmark -- --nocapture — 4/4 passing, BENCHMARK lines match benchmarks/baseline.json.
• cargo build -p iln_governance -p insurance_pool and cargo clippy --all-targets -p iln_governance — clean.
• docker compose config --quiet and docker compose -f docker-compose.test.yml config --quiet — both validate; docker compose config --services (no profile) confirms insurance-pool-deploy is excluded by default, and --profile insurance config --services confirms it's included when opted in.
• npx tsx scripts/gen-spec.ts / npx tsx scripts/gen-abi.ts — regenerate docs/contract-spec.json / docs/contract-abi.md idempotently (diff is empty on a second run).
• npx tsx --test scripts/generate-sdk.test.ts — 5/5 passing, confirming docs/contract-spec.json's top-level shape is unchanged.

[drips-wave]

@brightpixel-dev Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits