agentsclimarketplace

Project sprint roadmap

Skill Ink01101011/claude-mcp-kkskills/skills/project-sprint-roadmap

Use this skill whenever the trader-platform re-architecture comes up — Clean Architecture migration, 4-stage LLM pipeline, sprint planning, sprint scope, or any reference to Sprint 0–7+. Triggers include: mentions of 'sprint', 'roadmap', 'AUDIT', 'SCALING', 'BUG' tracks, 'Clean Arch', 'UnitOfWork'/'UoW', 'import-linter', 'entities/repositories/use_cases', 'chained pipeline', 'dual-write', 'strangler fig', 'aggregate' (Signal/Position/Analysis), 'PoC'. Also use when scoping new work — to decide which sprint it belongs in and whether it conflicts with existing scope. Do NOT use for unrelated codebases — this is trader-platform-specific.From its SKILL.md

Install
npx -y skills add Ink01101011/claude-mcp-kkskills --skill project-sprint-roadmap

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

6.4 KB, as published. Nobody here has run it

Project — Sprint Roadmap

Overview

The trader-platform re-architecture runs three tracks in parallel, not sequentially. Coordinated through 27052026_SPRINT_ROADMAP.md — the single source of truth.

Three tracks:

  1. 🐛 BUG (27052026_BUG.md) — stabilization, handed off to a different dev
  2. 🏗️ AUDIT (27052026_ARCHITECTURE_AUDIT.md) — Clean Architecture migration: entities/ + repositories/ + use_cases/ + UnitOfWork, with import-linter as the boundary fence
  3. 🧠 SCALING (27052026_SYSTEM_PROMPT_SCALING.md) — 4-stage chained LLM pipeline with dual-write rollout

When to Use

  • ✅ Any scope/planning conversation for trader-platform.
  • ✅ When deciding whether a piece of work is Sprint N or out-of-scope.
  • ✅ When pattern questions come up (entity? repo? UoW? import-linter contract?).
  • ❌ For ad-hoc bug fixes that have nothing to do with arch — those go through the BUG track or boy-scout rule.

Process / Steps

Step 1 — Look up the current sprint

  • ☐ Open docs/architecture/27052026_SPRINT_ROADMAP.md §2.
  • ☐ Identify which sprint is active and read its Goal, Tasks, Acceptance, Out-of-Scope.

Example: The user says "อยากเพิ่มเรื่อง X". You open §2, see the active sprint is Sprint 1 (Signal PoC). X is refactor analysis_engine.py — that's listed as Out-of-Scope for Sprint 1 (it's Sprint 3 work).

Step 2 — Place the work in a sprint

Use this lookup table:

If the work is...Sprint
Fix a critical bug from the bug list0 (BUG track)
New entities/, repositories/, use_cases/, UoW PoC1
Position aggregate, capability gating2
Analysis aggregate, chained 4-stage pipeline3
Dual-write rollout, disagreement reporting4
Promote new pipeline / hold / revert decision5
batch_jobs/ HTTP or gRPC split6
Migrate a services/ file when feature work touches it7+ (boy-scout)

Example: "Move portfolio_engine math out into its own module" → Sprint 2 (Position aggregate extraction).

Step 3 — Verify it doesn't violate Sprint 1's additive rule

  • ☐ New endpoint at a new path? Good.
  • ☐ Adds a column via ADD COLUMN ... NULL? Good.
  • ☐ Touches an existing endpoint's response shape? Stop — re-scope as additive.
  • ☐ Refactors analysis_engine.py? Stop — that's Sprint 3.
  • ☐ Splits batch_jobs/? Stop — that's Sprint 6.
  • ☐ Cleans up "while I'm here"? Stop — listed as an anti-pattern in §5.

Step 4 — Confirm dependency gates

  • ☐ Sprint 3 needs Sprint 2 done (capability gating must exist).
  • ☐ Sprint 4 needs Sprint 3 done (chained pipeline must exist).
  • ☐ Sprint 5 needs Sprint 4 + 14 trading days of dual-write data.
  • ☐ Sprint 6 needs Sprint 3 use cases (so HTTP split has a clean target).

Sprint Reference Table

SprintNameFocus
0Bug StabilizationBUG track. 3/7 critical done.
1Foundation WedgePoC with Signal aggregate, full-tree import-linter day 1.
2Position + capability gateAUDIT + SCALING contracts.
3The Big WedgeAnalysis + chained pipeline. AUDIT and SCALING overlap here.
4Dual-write rolloutSCALING.
5Promotion GateSCALING decision (wait for 14 trading days).
6batch_jobs/ HTTP splitAUDIT.
7+Boy-scout strangleAUDIT, no end date — migrate services/ only when feature work hits.

Definition of Sprint 1 success: "Sprint 2 dev opens the PR, reads CLAUDE.md + Signal example, starts Position aggregate without asking architectural questions, and import-linter shouts if they get the layer wrong."

Rules & Constraints

  • ALWAYS: when work is proposed, check it against the current sprint's Out-of-Scope list.
  • ALWAYS: Sprint 1 changes must be additive only.
  • ALWAYS: import-linter is full-tree from day 1 — legacy code can't add new violations, new code must comply from line 1.
  • NEVER: refactor analysis_engine.py before Sprint 3.
  • NEVER: split batch_jobs/ before Sprint 6.
  • NEVER: do "while I'm here" cleanup outside sprint scope.

Examples

Input: "อยาก refactor analysis_engine.py ตอนนี้เลย ได้ไหม" → Output: Pushback — that's Sprint 3. The pattern needs to be proven on Sprint 1 (Signal) and Sprint 2 (Position) first. Suggest a Sprint 1 PoC contribution instead.

Input: "Sprint นี้ใส่ feature X ได้ไหม" → Output: Cross-reference the current sprint's Goal and Out-of-Scope sections. If X isn't in scope, suggest deferring or routing through the BUG track if it's a real defect.

Input: "Clean Arch มี contract อะไรบ้าง" → Output: 5 import-linter contracts — entities-pure, repos-isolate-orm, use-cases-pure, routers-via-use-cases, no-backwards-strangle. Full config in 27052026_ARCHITECTURE_AUDIT.md §8.5.

Keep looking

Skills are one crate of 608,475. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.