Stats
Actions
Available In
Tags
Help us improve
Share bugs, ideas, or general feedback.
Plugin
Harness Base
Four non-negotiable safety hooks (command-guard, secret-scan, verify-gate, audit-log) plus cross-cutting engineering modules as auto-loading skills. Every Harness domain pack depends on this.
$
npx claudepluginhub camilool8/harness-engineering-templates --plugin harness-basePopularity
Stars
Med: 0·Avg: 274
Installs
Med: 0·Avg: 1
What's Inside
Agents8
architect-reviewer
/architect-reviewer
Pipeline stage 2 gate. Read-only. Reviews the spec for architectural soundness before any code is written. Different model family than the implementer.
implementer-worker
/implementer-worker
Executes ONE bounded task from the planner's plan in an isolated context (supervisor-worker topology). Has Edit/Write/Bash. Returns a typed diff summary. Invoke once per task.
implementer
/implementer
Pipeline stage 3. Implements an approved spec. Has Edit/Write/Bash. Returns a typed diff for the tester stage.
planner
/planner
Decomposes a goal into an ordered, typed plan of bounded tasks with acceptance criteria. Read-only — never edits code. Invoke once at the start of supervisor-worker work.
quality-reviewer
/quality-reviewer
Stage-two review. Read-only. Assesses code quality, maintainability and risk. Uses a DIFFERENT model family than the implementer to avoid sycophancy. Runs only after spec-reviewer passes.
Skills17
using-graph-memory
/memory-knowledge-graph
Store and query provenance-aware, time-bounded facts in the Zep knowledge-graph MCP. Use when working with facts that have a source, change over time, or need multi-hop reasoning across related entities.
managing-file-memory
/memory-md-files
Write, index and prune durable knowledge as markdown notes in .claude/memory/. Use when you learn a fact, make a decision, resolve an incident, or discover a recurring procedure that a future session would need.
using-vector-memory
/memory-vector-store
Store and retrieve durable facts from the Mem0 vector-memory MCP. Use at session start to recall task-relevant memories, and whenever you learn a durable user- or domain-fact worth keeping.
authoring-feature-files
/methodology-bdd
Writes Gherkin .feature files from a user story so non-technical stakeholders can sign off on behavior. Use before implementing any user-visible behavior change.
running-evals
/methodology-eval-driven
Builds and iterates evals for LLM/ML output. Use whenever changing a prompt, model, retrieval config, or tool surface, or when an output's correctness is judgmental rather than deterministic. Arm the Stop gate with eval = true in .claude/HARNESS.toml.
Stats
Version0.1.0
LanguageShell
Stars0
MaintenanceGood
LicenseMIT
Last Commit
Added
Available In
Safety Signals
Critical
Matches all tools
Hooks run on every tool call, not just specific ones
Caution
Executes bash commands
Hook triggers when Bash tool is used
Help us improve
Share bugs, ideas, or general feedback.
