19 pages tagged

#claude-code

agent-legibility

Designing systems for the agent's eye, not the human's; logs, metrics, code all agent-readable.

architectural-invariants

Strict layering enforced by custom linters with embedded fix instructions.

Lifecycle shell hooks; deterministic enforcement of things you don't trust the model with.

The project-root contract file; short, hard, executable — not a wiki.

On-demand knowledge/workflow packages; descriptors resident, bodies progressively disclosed.

claude-subagents

Spawned Claude instances with isolated context; about isolation, not parallelism.

context-engineering

Choosing what enters the LLM context; steering-not-teaching; tiers, Context Rot, compaction.

loop-engineering

Anthropic's four-type loop taxonomy (turn/goal/time/proactive); the hand-off ladder.

model-and-effort-selection

Claude Code's two dials: model = capability, effort = thoroughness; triage + token economics.

model-context-protocol

MCP: how external systems plug into Claude Code; largest hidden context cost.

Read-only exploration mode; refines a plan before any file is touched.

Anthropic's prefix-cache mechanism; the architectural backbone Claude Code is shaped around.

six-layer-agent-architecture

Tw93's Claude Code frame: CLAUDE.md / Tools+MCP / Skills / Hooks / Subagents / Verifiers.

The closing layer; "Claude said done" isn't done — bind acceptance criteria up front.

Anthropic's CLI / agent harness; loop primitives + model/effort dials.

Alibaba/Fliggy engineer; author of all three seed essays and of [[kaku]].

2026-04-27-claude-code-architecture-governance-engineering

Tw93's six-layer frame for Claude Code: context, tools, Skills, Hooks, Subagents, verifiers.

2026-07-12-claude-model-effort-level

Anthropic's model-vs-effort explainer: "didn't know vs didn't try" triage.

2026-07-12-loop-engineering-getting-started

Claude Code team's loop taxonomy: turn/goal/time-based + proactive; stop conditions.