agentic-development-principles

Universal principles for agentic development when collaborating with AI agents. Defines divide-and-conquer, context management, abstraction level selection, and an automation philosophy. Applicable to all AI coding tools.

Install with Tessl CLI

npx tessl i github:supercent-io/skills-template --skill agentic-development-principles

What are skills?

1.01x

Quality

47%

Does it follow best practices?

Impact

89%

1.01x

Average score across 3 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./.agent-skills/agentic-development-principles/SKILL.md

SKILL.md

Review

Evals

Evaluation results

85%

-1%

Webhook Notification Service: Implementation Plan

Staged task decomposition

Criteria

Without context

With context

Multiple distinct steps

100%

Stage 1: Design/schema

100%

Stage 2: Core logic

87%

100%

Stage 3: API/interface connection

100%

Stage 4: Tests

40%

20%

Stage 5: Integration/refactor

62%

25%

Independent validation

100%

Scoped steps

80%

100%

Concrete deliverable per step

100%

Fix-in-isolation design

80%

90%

Non-monolithic start

100%

Without context: $0.3259 · 1m 53s · 15 turns · 21 in / 5,603 out tokens

With context: $0.3138 · 1m 58s · 12 turns · 452 in / 5,761 out tokens

100%

Session Transition: Backend Authentication Work

Context handoff documentation

Criteria

Without context

With context

HANDOFF.md created

100%

Completed work section

100%

Current status section

100%

Next tasks section

100%

Tried but failed section

100%

Cautions section

100%

Single-topic focus

100%

Concise essentials only

100%

Failed attempts include reason

100%

Current status specificity

100%

Without context: $0.1059 · 32s · 8 turns · 13 in / 1,656 out tokens

With context: $0.2397 · 49s · 12 turns · 197 in / 2,338 out tokens

83%

Database Layer Refactor: Migrate from Raw SQL to Query Builder

Plan-first execution and verification

Criteria

Without context

With context

Plan document created

100%

Risky operations identified

30%

Backup strategy mentioned

Staged execution

70%

100%

Tests written

100%

Edge cases in tests

100%

Self-verification summary

100%

Diff/change review

100%

No dangerous commands unguarded

100%

Behavior preservation check

100%

Without context: $0.4652 · 2m 8s · 23 turns · 27 in / 7,984 out tokens

With context: $1.2911 · 4m 27s · 47 turns · 490 in / 16,812 out tokens

Evaluated: 3 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Webhook Notification Service: Implementation Plan Session Transition: Backend Authentication Work Database Layer Refactor: Migrate from Raw SQL to Query Builder

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.