agentleFS
neilyongluneilyonglu/project-development/SKILL.md

project-development

Skill0 starsChanged 38 days agoMITView on GitHub

Integrate project-specific architecture, conventions and tool dependencies, then guide requirements analysis, design, acceptance, planning, implementation and two-pass review. Use for project-development workflows or a requested stage; keep routine small fixes scoped. Supports English and Traditional Chinese deliverables.

11 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use self-improvement-loops, long-horizon-prompting and 3 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • DVwebBiz11Skill

    self-improvement-loops

    This skill should be used when the harness, scaffold, workflow, or optimizer itself is the optimization target: recursive self-improvement (RSI) loops, meta-harnesses, self-improving harnesses that mine their own failures and propose bounded edits, evolutionary or population-based search over agent scaffolds, acceptance gates for self-modifying systems, and agentic context evolution where the mechanism that produces context is versioned and evolved. Route governance of a single autonomous loop (locked surfaces, durable logs, rollback, novelty gates, approval boundaries) to harness-engineering, measurement and quality-gate design to evaluation, judge design to advanced-evaluation, and remote sandbox infrastructure to hosted-agents.

    Also in 4 projects that use this

  • DVwebBiz11Skill

    long-horizon-prompting

    This skill should be used when writing, enhancing, or evaluating the launch prompt for a long-running autonomous agent or a parallel multi-agent orchestration attacking a hard problem: pseudo-formal task briefs that define terms and an exact success predicate linguistically, enumerate non-counting outcomes, set persistence rules with explicit stop and return conditions and effort floors, manage a diverse portfolio of parallel approaches with an approach registry and blocked-route bookkeeping, and gate the return on adversarial audit. Route agent topology and coordination protocols to multi-agent-patterns, runtime control surfaces and loop governance to harness-engineering, evaluator and quality-gate construction to evaluation, judge design to advanced-evaluation, and compaction or memory mechanics to context-compression and memory-systems.

    Also in 4 projects that use this

  • AbdelrhmanAlquranSkill

    latent-briefing

    This skill should be used when the user asks to "share memory between agents", "KV cache compaction for multi-agent", "orchestrator worker context", "latent briefing", "reduce worker tokens", "cross-agent memory without summarization", or discusses Attention Matching compaction, recursive language models with workers, or token explosion in hierarchical agents.

    Also in 4 projects that use this

  • GitHubSkill

    harness-engineering

    Adopt repository-level harness engineering for coding agents. Use when a user wants to prevent repeated AI coding-agent mistakes by turning failures into durable instructions, drift checks, regression tests, failure memory, and adoption reports tailored to the target repository.

    Also in 5 projects that use this

  • BlockSkill

    frontend-design

    Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the requested stages of the author's development workflow.

    line 12
  2. Very specific:

    Use this English entrypoint and the English guides under references/ as the shared operating instructions for both output modes.

    line 16
  3. Distinctive:

    Follow the English operating instructions for analysis and checks.

    line 19
  4. Very specific:

    Use Chinese HTML template for zh-TW and English HTML template for en.

    line 22
  5. Distinctive:

    Add further delivery stages only when needed and confirmed.

    line 46
  6. Distinctive:

    Document units follow project context.

    line 57

“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “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

Project Development

Use the requested stages of the author's development workflow. Read only the stage guidance and dependencies needed for the current task.

Language modes

Use this English entrypoint and the English guides under references/ as the shared operating instructions for both output modes. Traditional Chinese guide and references/zh-TW/ are human-readable translations, not a separately installed skill. Do not load both editions by default.

  • Resolve output_language from the user's explicit choice for the current work, then the project's recorded preference, then the default zh-TW. Supported modes are Traditional Chinese (zh-TW) and English (en). Carry an explicit selection forward until the user changes it; support another requested language without overriding the user.
  • Follow the English operating instructions for analysis and checks. This specifies instruction language, not a guarantee about private reasoning language. Do not request or expose private chain of thought; deliver conclusions, evidence, assumptions and verification results.
  • Apply the selected language to replies, profile summaries, analysis/design/planning documents, review records, diagram explanations and visible HTML controls. Preserve code identifiers, API/DB fields, commands, paths and verbatim source quotations. Code comments follow the project's conventions unless the user requests translation.
  • Record an explicitly chosen persistent project preference in its context document. Do not infer a project-wide preference from one translated quotation. Do not translate all existing project documents or create a second work document just because the output mode changes.
  • Use Chinese HTML template for zh-TW and English HTML template for en; they share the same data schema and behavior. Replace example content in the selected language.
  • Maintain both documentation editions when changing workflow requirements. English is the maintained execution edition; resolve translation discrepancies against it without introducing a different workflow.

Examples: "Use project-development with English output"; "使用 project-development,輸出用繁體中文".

Integrate project context first

On first entry, follow project context to create or refresh a project-local profile: architecture, module conventions, data and interfaces, requirements/document units, development and delivery flow, verification and tool dependencies. Subsequent stages use that profile and its primary sources to preserve the user's development logic.

Reuse an existing profile and check only relevant rules and changes. Context integration permits reading and documentation, not product construction or external actions. Missing unrelated information does not block the requested stage.

Stages

StageWhen to useGuideDeliverable
Issue / requirements analysisClarify requirements, intended architecture and implementation impactissue-analysisFive-part analysis under the project's document unit
Project scaffoldingUnderstand the framework and DB design or build an agreed skeletonproject-scaffoldFramework/table design; skeleton, config, migration and basic verification when construction is requested
Operations and data flowExplain how an operation connects requirements, modules and DBoperation-flowEditable flow/sequence diagrams, reads/writes, states, failure branches and open questions
Interface contractsConfirm cross-module inputs, outputs, errors and UI Props mappingsinterface-contractField/validation tables, result/error contracts, UI mappings and discrepancies
Business acceptanceConfirm expected behavior for data, operations and boundariesbusiness-acceptanceSource-backed concrete acceptance cases, data changes and unresolved questions
Implementation planningTurn confirmed design and acceptance into executable workimplementation-planGaps, tasks, dependencies, completion conditions and first step
Implementation and validationImplement the authorized scope and verify acceptance behaviorimplementation-validationCode changes, verification evidence, limitations and remaining work
Review and deliveryClean up, perform two review passes and complete architecture/test mappingreview-deliveryReview/verification record and implementation/test HTML view

Add further delivery stages only when needed and confirmed; do not create empty workflows or assumed policies.

Build implementation diagrams, including file walkthrough progress, from the selected language template using architecture diagram. For repeated inventory, coverage, structure and local-link checks, use automation when needed; scripts cannot replace requirements or semantic code review.

Selection and handoffs

  • Analysis-only requests use issue-analysis; construction requests use project-scaffold with sufficient requirements/architecture evidence.
  • Continue across stages within existing authorization. Completing one stage does not independently authorize the next, and prior authorization does not need to be requested again.
  • Requests to understand or discuss a design stay in explanation/proposal mode. Do not infer construction permission from a discussion or preview.
  • Unresolved choices affecting the framework, core model, permissions or writes block their dependent work, not independent work.
  • Reuse the real framework and upgrade mechanism in existing projects. Evaluate frameworks for new projects; this skill does not imply rebuilding, adding tables or expanding scope.
  • Document units follow project context. An issue-based project may use one document per issue, updated throughout the work. A guide's "issue/work document" means the current project's chosen unit; do not impose another project's platform, numbering or directory.
  • Reuse confirmed unchanged design; do not reanalyze the whole issue or demand every preceding artifact at each stage.
  • Maintain requirements → design/tasks → implementation locations → acceptance cases/results in the original work document. Each in-scope requirement has a destination; each change has a requirement or necessary technical basis. Address omissions and scope drift without inventing requirements.
  • Shared checks are defined in the relevant guide and referenced elsewhere. Keep responsibility, authorization and links consistent. Design confirmed, code written, tests passed and ready to publish are different states.

Examples: use issue-analysis for #101; use project-scaffold to build the modules and migration files needed by confirmed requirements.

Locate this skill's scripts and assets from its actual loaded directory, not a fixed .claude/skills path. See README for packaging and evaluation.

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.