agentleFS
YuY-QKYuY-QK/agent-kit/skills/architecture-change/SKILL.md

architecture-change

Skill0 starsChanged 22 days agoMITView on GitHub

Design or review cross-module architecture, dependency direction, databases, file formats, networking and integrations, long-running tasks, payments, permissions and security boundaries. Use it for structural changes; don't use it for local UI tweaks or simple function edits.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. florian101010/awesome-agentic-AI-coding-template★ 4same name
    2. accnet/claude-skill★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository7 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the document routing in the project's AGENTS.md to find the technical design, ADRs and data or interface specifications, and read only …

    line 12
  2. Distinctive:

    When interfaces, file formats, databases, permissions or release boundaries change, also read and maintain the matching contracts, …

    line 13
  3. Distinctive:

    When product scope or entitlements change, check the PRD.

    line 14
  4. Distinctive:

    When there is a real trade-off, compare the current approach with at least one credible alternative.

    line 23
  5. Distinctive:

    Prefer a single source of truth, local changeability and clear boundaries.

    line 24
  6. Distinctive:

    Don't build general platforms, plugin systems or abstraction layers for features that haven't been approved.

    line 28
  7. Distinctive:

    Before integrating real payments, confirm the product and price source, the purchase state machine, receipt validation or the server …

    line 36

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Architecture and boundary changes

Give structural changes a design that can evolve, migrate, roll back and be verified — without over-engineering.

Read the material

  • Use the document routing in the project's AGENTS.md to find the technical design, ADRs and data or interface specifications, and read only the sections that affect this change.
  • When interfaces, file formats, databases, permissions or release boundaries change, also read and maintain the matching contracts, migration notes and release checklists.
  • When product scope or entitlements change, check the PRD; if there is no product document yet, ask the user to confirm the key contracts first.
  • Documents describe the target; code and tests prove the current state. Don't mix them up.

Define the boundaries first

Make explicit: the problem, constraints, non-goals and quality requirements; the authoritative data source, derived data, owners and lifecycle; module responsibilities, dependency direction, inputs/outputs and how failures propagate; sync/async, concurrency, cancellation, retries, idempotency and atomic commits; compatibility, migration, rollback, recovery and observability; and the impact of privacy, security, offline use and external-service failures.

Design principles

  • When there is a real trade-off, compare the current approach with at least one credible alternative; don't force a comparison for a small change with only one sensible option.
  • Prefer a single source of truth, local changeability and clear boundaries; the number of layers or interfaces is not a measure of quality.
  • The UI layer never talks to databases, files, HTTP, payment SDKs or secure storage directly.
  • Core local functionality doesn't depend on the network; remote failures must not damage data already saved locally.
  • Development mocks and real data sources are swapped behind the same repository/data-source boundary.
  • Don't build general platforms, plugin systems or abstraction layers for features that haven't been approved.
  • Designs that break data or protocol compatibility need explicit approval first, plus versioning, migration, rollback and tests.
  • Record important decisions the user signs off on as ADRs.

High-risk gates

Changes to databases, file formats, caches, backups, payments, licences, keys and platform permissions must cover: reading and upgrading old versions and upgrade failures; interruption, repeated runs and partial success; corrupted, forged, out-of-range, oversized input and full disks; staging-area validation, atomic commit and cleanup on failure; and redaction in logs.

Before integrating real payments, confirm the product and price source, the purchase state machine, receipt validation or the server boundary, restore purchases and idempotent fulfilment; until then, never implement a fake success.

Output

Describe the background, constraints, current structure, options, recommended decision and its cost, modules and data flow, failure semantics, migration and rollback, testing strategy, documentation impact and open questions — only the parts relevant to this change. When the user only asked for a design or review, don't implement it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.