[Hackathon] cloud-post-code-pareto-negotiation: multi-attribute bargaining, Pareto frontier, agents trading off price vs deadline

[cloud-post-code]

What this PR does

The default alternating_offers plugin bargains over price only. This PR adds a second plugin, pareto, where agents bargain simultaneously over price, deadline, and quality and converge toward Pareto-optimal agreements: deals where neither side could be made better off without making the other worse.

All changes are scoped strictly to the negotiation layer, the multi_attribute_market scenario, validator additions, and leaderboard/docs. No shared infrastructure files (sim/, metrics.py, marketplace.py, alternating_offers.py) were modified.

---

Spec deliverables (§07 — Multi-attribute negotiation with Pareto-frontier search)

Criterion	Where
Plugin registered as ("negotiation", "pareto")	plugins.py +1 line
Bargains over ≥ 2 attributes from Terms	price + deadline (n) + quality tier
Each agent has a private utility function	ParetoParams per agent, never shared
Utility weights passed via constructor only	ParetoNegotiator(agent_id, params)
respond() evaluates full Terms	checks p, n, q feasibility + utility
Converges toward Pareto-optimal agreements	fewer dominated agreements than alternating_offers baseline
Adversarial validator FAILs alternating_offers	10/10 sessions quantity-dominated (upstream alternating_offers unchanged)
Adversarial validator PASSes pareto	fewer violations than baseline in every run
Scenario with 50 client-consultant pairs	scenarios/multi_attribute_market.yaml
Deterministic params from seeded RNG	hash(agent_id) ^ seed via _agent_rng()
Does not break alternating_offers	plugin extended, not replaced; alternating_offers.py untouched
Breakdown sessions labeled and excluded	pairs 7 & 8 emit close:breakdown:
Leaderboard entry	rank #2: 80% deal rate, 69.6 msg/tick throughput, 100% delivery

---

Attribute model

price.amount              → p   (int 0–100)    distributive axis — buyer wants low, seller wants high
conditions["deadline"]    → n   (int ticks)    integrative axis  — sweet-spots differ, joint optimum reachable
conditions["quality"]     → q   (float 0–1)   static per session; gates feasibility only

---

Six strategies (spec equations 10–12)

Strategy	Concession rule
ttt_directional	Mirror a bounded fraction of the opponent's per-issue movement on each axis
zeuthen_concede	Concede when own Zeuthen risk ratio ≤ opponent's; decrement tau by tau_step
rubinstein_discount	Accept when u_i(x_j) ≥ delta^t · V_i; discount continuation value each round
logrolling_tradeoff	Stay on own iso-utility contour; maximise estimated opponent utility
nash_bargaining	Asymmetric Nash product (u_i − d_i)^β · û_j^(1−β) over own feasible grid
ks_bargaining	Kalai-Smorodinsky: minimise `

Strategies assigned deterministically: buyer for pair i gets STRATEGY_IDS[(i*2) % 6], seller gets STRATEGY_IDS[(i*2+1) % 6]. Buyer and seller always get different strategies.

---

Files changed

New / substantially expanded

File	Net lines	What it does
packages/nest-plugins-reference/nest_plugins_reference/negotiation/pareto.py	+1,463	ParetoParams, utility model, build_feasible_grid, pareto_nondominated, 6 strategy classes, ParetoNegotiator
packages/nest-core/nest_core/scenarios_builtin/multi_attribute_market.py	+741/−576	50 client-consultant pairs; joint-zone opening; clamped counters; breakdown pairs 7 & 8
packages/nest-core/nest_core/validators.py	+374	_parse_pnq(), _pair_key(), Pareto-dominance checker, adversarial validator
packages/nest-plugins-reference/tests/test_pareto_negotiation.py	+615	8 integration tests: determinism, frontier, acceptance, breakdown, validator pass/fail
packages/nest-plugins-reference/tests/test_pareto_strategies.py	+634	All 6 strategies self-play + all 36 cross-strategy pairs
packages/nest-plugins-reference/tests/test_pareto_properties.py	+256	Hypothesis: monotonic concession, individual rationality, determinism
packages/nest-core/tests/test_metrics.py	+233	Bilateral negotiation metrics, drop_rate, self-message exclusion

Small modifications

File	Change
packages/nest-core/nest_core/plugins.py	+1 line: registers ("negotiation", "pareto")
scenarios/multi_attribute_market.yaml	Scaled to 50 pairs; task.type → multi_attribute_market
docs/layers/negotiation.md	+95 lines: attribute model, strategy catalog, validator logic, trace protocol
apps/nest-dashboard/src/lib/demo-data.ts	+17: rank #2 entry (80% deal rate, 69.6 msg/tick, 100% delivery)
apps/nest-dashboard/src/app/leaderboard/page.tsx	+3: renders dealRate column

Explicitly NOT modified (confirmed upstream)

• sim/transport.py, sim/simulator.py, metrics.py, marketplace.py, alternating_offers.py

---

What was fixed vs the prior submission (hackathon/The_Dealmakers)

Reviewer issue	Resolution
Branch name didn't match charter convention	New branch: hackathon/cloud-post-code-pareto-negotiation
Cross-session learning out of scope (§07)	dealmakers.py, test_dealmakers.py, dealmakers.yaml, scenarios.py registration — all removed
sim/transport.py global semantic change	Reverted to upstream
sim/simulator.py self-message bypass	Reverted to upstream
metrics.py redefines shared scoring metrics	Reverted to upstream
marketplace.py shared builtin rewritten	Reverted to upstream
alternating_offers.py baseline changed	Reverted to upstream; validator tests against unmodified reference
marketplace_failures.yaml, multi_attribute_market_altoffers.yaml out of scope	Removed
PR description omitted ~1,700 lines from files-changed table	Full accurate manifest above

---

🤖 Generated with Claude Code