We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Correct the sensor claims, add References, document what the eval score is not
Add Agent Pipelines page with one sequence diagram per agent
docs(v5): flip status to live across all stages, link canonical pattern spec intake + resolve-mark-proceed + adversarial eval now rolled out to prd, prp, plan, dev, test, pr, and system-design. Point pages at .claude/shared/pipeline-pattern.md. Only Haiku tiering + per-org repos.md remain (planned). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
docs(v5): add Autonomy + Orchestration-and-Subagents, link from Home/Sidebar/Agents/Pipeline Document the v5 direction: intake stage, resolve-mark-proceed, gated autonomy (humans on the loop), and orchestrator + leaf-subagent topology with the adversarial writer/critic split. Mechanisms marked live or (planned) as the prd-first rollout proceeds. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Designer Skill: harden iconography (no emoji, original modern icons)
Add Designer Skill page; document it on Agents, Home, Sidebar
Remove em-dashes across all wiki pages
Expand wiki into full harness-kit knowledge base: harness engineering, guides, sensors, evals, pipeline, golden path, agents
Merge branch 'master' of https://github.com/Pierry/harness-kit.wiki
Initial Home page
Add System Architect system design wiki: method + URL shortener, rate limiter, search engine