claudecodex

harness-self-improvement

SIPS graph runtime and home-base for bounded task execution, memory fabric recall, verification, delegation, and lifecycle automation.

0 on GitHubby RasputinKaiserv0.4.0MITupdated 1 month agodevelopment
/

Plugin harness-self-improvement from marketplace sips-local in RasputinKaiser/Hermes-SIPS. Commands read from the repo's manifest — 1 month ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.

What it installs

Skills (10)

  • sips-context-distiller

    Extract bounded, source-linked context from large files or many inputs. Use when context is too large, files are oversized, or a task needs concise excerpts.

  • sips-control-plane

    Inspect SIPS Homebase status, manifest wiring, routes, host visibility, and MCP freshness. Use when asked for SIPS status, command center, host audit, plugin visibility, or whether Homebase is fresh.

  • sips-delegation-router

    Route a task through SIPS commands, agents, scripts, MCP tools, or bounded delegation. Use when asked to route, split, fan out, escalate, or choose the right SIPS path.

  • sips-execution-repro

    Turn failures, logs, symptoms, or flaky behavior into a compact repro and verification plan. Use when debugging needs a reproducible failure path.

  • sips-memory-fabric

    Search, inspect, and record SIPS-owned Memory Fabric lessons. Use when a task needs prior lessons, recurring-fix memory, recall health, scoped historical context, or when a just-fixed bump or error should be recorded.

  • sips-perception-plan

    Plan browser, app, screenshot, or UI checks before visual claims. Use when a task includes app shots, visual QA, generated assets, or runtime UI proof.

  • sips-proof-scanner

    Verify SIPS proof surfaces before reporting completion. Use when asked to scan, verify, validate, prove, audit, or check whether a repo or SIPS install is coherent.

  • sips-repo-map

    Map repository structure, likely tests, git state, and write scope before edits. Use when asked for repo mapping, blast radius, file inventory, or change planning.

  • sips-selfloop

    Start, inspect, or continue a persistent SIPS self-improvement loop. Use when the user asks for /selfloop, wants the agent to iteratively improve itself, or wants a goal dedicated only to SIPS and agent capability.

  • sips-tool-factory

    Decide whether to reuse, improve, or scaffold a deterministic helper. Use when repeated work is slow, brittle, or blocked by a missing local tool.

Slash commands (12)

  • /harness-self-improvement:brainstorm

    Survey current capabilities, identify gaps, dispatch the escalate agent to draft a concrete build plan for the top gap. Proactive — the complement to /improve (which is reactive).

  • /harness-self-improvement:checkpoint

    Snapshot the harness and write a continuity packet. Safe point before risky work.

  • /harness-self-improvement:escalate

    Force-route the next bounded step to the frontier (claude-opus) escalation agent.

  • /harness-self-improvement:fan-out

    Decompose a parent task into parallel slices and dispatch one fan-out agent per slice. Each slice runs in its own slice directory with a HANDOFF.md describing the parent and its scope. Use /fan-out "parent objective" with slice descriptions as args (semicolon-separated).

  • /harness-self-improvement:goal

    Set a goal and enter the RALPH loop — keep working autonomously until the goal is genuinely done or the user stops you. Use /goal "objective" to start, /goal status to check, /goal complete to finish, /goal clear to stop.

  • /harness-self-improvement:improve

    Run a self-correction sweep and act on the top recommendation. Closes the loop on demand.

  • /harness-self-improvement:patterns

    Show the full agent_patterns report — success rate, approach→outcome correlation, top patterns.

  • /harness-self-improvement:recall

    Tier-aware Memory Fabric search for the current scope. Prints ranked prior lessons.

  • /harness-self-improvement:retro

    Run an evidence-gated retrospective, capture reusable SIPS lessons with provenance, and leave one clear next action.

  • /harness-self-improvement:selfloop

    Start or control a persistent SIPS loop whose only objective is improving SIPS and the agent operating it through measured, verified iterations.

  • /harness-self-improvement:teach

    Hand-record a lesson into Memory Fabric (learning tier, high confidence) so it is recalled later.

  • /harness-self-improvement:verify

    Run run_tests.py + script_smoke on touched files. If a coverage gap is found, offer test-author.

Subagents (5)

  • escalate

    Solve one bounded subtask the main thread is stuck on, in a fresh context. Returns a plan, a minimal diff, and a one-line lesson to record.

  • fan-out

    Solve one slice of a decomposable task independently. Other fan-out agents may be working on other slices in parallel. Coordinate via the shared HANDOFF.md file the main thread wrote; return your slice, a minimal diff, and a one-line lesson.

  • memory-curator

    Dedupe near-identical Memory Fabric records, promote repeated work-tier entries to learning-tier, and expire stale low-confidence ones. Keeps recall signal high.

  • repo-scout

    Cheap parallel codebase recon — find files, map call sites, summarize a module. Fan many of these.

  • test-author

    Given an untested script, write the missing run_tests.py regression case that would have caught a known failure or covers its public surface.

MCP servers (1)

  • sips-homebase

    runs: python3 ./scripts/harness_homebase_mcp.py

Hooks (20)

  • PreToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PreToolUse --script autonomy_gate.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/autonomy_gate.py

  • PreToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PreToolUse --script memory_fabric_preflight.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/memory_fabri

  • PreToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PreToolUse --script autonomy_gate.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/autonomy_gate.py

  • PostToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostToolUse --script script_smoke.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/script_smoke.py

  • PostToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostToolUse --script escalation_advisor.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/escalation_advis

  • PostToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostToolUse --script sips_presence_mirror.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/sips_presence_

  • PostToolUse

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostToolUse --script retry_lesson_reminder.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/retry_lesson_

  • SessionStart

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event SessionStart --script validate_harness.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/validate_harness.

  • SessionStart

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event SessionStart --script memory_fabric_doctor.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/memory_fabric

  • SessionStart

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event SessionStart --script proactive_drift.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/proactive_drift.py

  • SessionStart

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event SessionStart --script agent_patterns.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/agent_patterns.py -

  • SessionStart

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event SessionStart --script improvement_injector.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/improvement_i

  • UserPromptSubmit

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event UserPromptSubmit --script recall_ranker.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/recall_ranker.py

  • UserPromptSubmit

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event UserPromptSubmit --script probe_hook.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/probe_hook.py

  • PreCompact

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PreCompact --script memory_fabric_compact_brief.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/memory_f

  • PreCompact

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PreCompact --script compact_continuity.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/compact_continuit

  • PostCompact

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostCompact --script memory_fabric_session_record.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/memory

  • PostCompact

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event PostCompact --script compact_continuity.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/compact_continui

  • Stop

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event Stop --script session_close.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/session_close.py

  • Stop

    python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/hook_event_tap.py --event Stop --script task_outcome_tracker.py -- python3 "${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}"/scripts/task_outcome_tracker.

View source on GitHub

Similar plugins

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

context7

claudecodex

Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.

1 MCP server

upstash/context7 64,709

Build powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.

5 commands

HKUDS/CLI-Anything 58,665

Automates browser interactions for web testing, form filling, screenshots, and data extraction

12 skills

vercel-labs/agent-browser 50,373

Use Alibaba Open Code Review from local Codex.

2 skills

alibaba/open-code-review 50,068

slidev

claude

Claude Code skill for creating and presenting Slidev slidedecks

1 skill

slidevjs/slidev 49,452

Maintain harness-self-improvement?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted