lo2cin4bt-acceptance

name	lo2cin4bt-acceptance
description	Repo-local acceptance skill for lo2cin4bt. Use when checking whether a deliverable satisfies the user request, repo contracts, skills, public/GitHub boundary, tests, stale docs, forbidden paths, and no-trading-advice disclaimers.
version	2.2.1
status	active
category	auditor
use_when	["Checking a lo2cin4bt deliverable against user requirements and current contracts.","Reviewing architecture drift, stale docs, output boundaries, or public claims."]

lo2cin4bt Acceptance Skill

Purpose

Provide skill-owned acceptance for user intent, shared Rust architecture, artifacts, frontend contracts, safety, and evidence.

Use When

Use before closing meaningful lo2cin4bt code, config, agent, skill, Lecture, frontend, or release work.

Inputs

• User request and latest corrections.
• Changed files, current contracts, tests, payloads, artifacts, and reports.

Workflow

Apply the Checklist below, run risk-matched validation, return pass/revise/block, and reject structural drift rather than accepting exceptions.

Language Policy

1. Use response_language from the PM packet when present; otherwise infer it from the latest user message.
2. If the user writes in Chinese or asks for Chinese, write Traditional Chinese for all non-specialist wording.
3. Keep code identifiers, file paths, commands, schema keys, agent/skill names, ticker symbols, provider names, and standard finance/quant abbreviations exact. For mixed technical terms, write Chinese first with English in parentheses where useful, e.g. 夏普率 (Sharpe), 前向分析 (WFA).
4. Return response_language in the output packet. When writing repo-local reports, apply the same language policy to the report body.

Safety Notice

Acceptance review checks deliverable quality only. It does not endorse any strategy and is not investment advice, trading advice, financial advice, or an instruction to trade.

Required Reads

• The user request and latest corrections
• skills/lo2cin4bt/references/runtime-architecture.md
• The agent and skill contracts touched by the task
• Relevant README/doc/config/test artifacts
• skills/lo2cin4bt/references/readme-acceptance-criteria.md when README or public docs are touched
• skills/lo2cin4bt/references/lo2cin4-agent-contract.md when AI agent discoverability is touched
• skills/lo2cin4bt/references/workspace-and-github-boundary.md when GitHub/public boundaries are touched
• skills/lo2cin4bt/references/strategy-identity-and-summary.md when strategy config, selector, summary, or result payloads are touched
• Public GitHub boundary docs when release readiness is involved

Checklist

• user wording followed
• latest correction supersedes older instructions
• no scope drift
• required agents/skills/references exist
• no unsupported strategy/config claim
• no public performance overclaim
• no live trading, broker action, external account action, or production deployment
• docs and skills agree on paths and names
• evidence is current
• tests or scans match change risk
• new strategy configs use a specific four-part platform.display_label with no Backtest, workflow, date, or run id prefix
• result-selector and strategy-logic previews are derived from the executable config, not copied from user prose or metadata.notes
• app-managed run identity and Strategy Logic agree with the saved strategy_run.json snapshot
• Python remains control-plane only and all result-changing calculations use the shared Rust engine, mandatory Rust validator, Rust metrics, and PlotBundle contract
• WFA and rolling validation remain explicit validation_workflow/ workflows; Parameter Matrix is not mislabeled as validation
• plotter/web is the only frontend and the app serves it on port 2424
• the resolved Git root is the product Repo directory itself
• product origin fetch/push URLs exactly match the approved GitHub repository
• product main is clean, contains the fetched remote history, and is neither behind nor diverged from origin/main
• when nested in Company, the parent tracks only the product commit as a Git submodule, has no product remote, and never pushes product code
• candidate and tracked-index release guards pass; otherwise publishing is blocked without exceptions

Verdicts

• pass
• revise
• block

Output Format

response_language:
acceptance_verdict:
requirements_checked:
evidence_checked:
gaps:
scope_drift:
forbidden_or_out_of_scope_items:
required_followup_gate:
repo_agent_report_path:
not_trading_advice_notice:

Outputs

• Acceptance verdict with requirements, evidence, gaps, blockers, and required follow-up.

Path Rules

• Inspect product evidence under the current <repo-root>.
• Write acceptance evidence to the PM-assigned repo-local or Company report root.
• Do not move ignored runtime outputs into Git-tracked source paths.

Validation

uv run --locked --exact --group dev python -m pytest tests/test_agent_skill_lecture_alignment.py tests/test_strategy_run_config.py tests/test_app_api_payloads.py -q

Pass criteria: user requirements and current architecture agree, selected tests pass, and no unsupported claim or retired path remains.

Fail action: return revise or block with exact evidence; never waive a result-integrity or safety failure.

Report Requirements

Record requirements checked, evidence paths, commands, verdict, scope drift, residual risks, and the next required owner.