@neikyun/ciel 6.0.6 → 6.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/assets/platforms/opencode/.opencode/agents/ciel-critic.md +0 -1
- package/assets/platforms/opencode/.opencode/agents/ciel-explorer.md +0 -1
- package/assets/platforms/opencode/.opencode/agents/ciel-improver.md +0 -1
- package/assets/platforms/opencode/.opencode/agents/ciel-researcher.md +0 -1
- package/package.json +1 -1
|
@@ -1,7 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
description: Isolated-context critic subagent for Ciel. Dispatch when the main session needs hostile review (RELIRE), full 7-step audit (CRITIQUER), or root-cause analysis (RCA). Three modes — MODE=RELIRE (3 RISQUE after write), MODE=CRITIQUER (post-hoc audit), MODE=RCA (debug root cause). Always use for Critical tasks. Fresh context prevents degeneration-of-thought (CriticBench 2024). Tools — read/grep/bash allowed, edit/write denied.
|
|
3
3
|
mode: subagent
|
|
4
|
-
model: anthropic/claude-sonnet-4-6
|
|
5
4
|
temperature: 0.2
|
|
6
5
|
tools:
|
|
7
6
|
write: false
|
|
@@ -1,7 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
description: Isolated-context explorer subagent for Ciel. Dispatch for CODEBASE + FLUX steps — pattern-fitness-check, flux-narrator, domain mastery, modern-patterns-checker, ai-failure-modes-detector, test-strategy, playwright-visual-critic, cicd-security-hardener, accessibility-wcag-auditor. Reads the codebase fresh, free of main-session bias. Tools — read/grep/glob allowed, no bash/edit/write.
|
|
3
3
|
mode: subagent
|
|
4
|
-
model: anthropic/claude-haiku-4-5-20251001
|
|
5
4
|
temperature: 0.2
|
|
6
5
|
tools:
|
|
7
6
|
write: false
|
|
@@ -1,7 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
description: Long-running meta-agent for Ciel self-improvement. Dispatch ONLY on /ciel-improve, /ciel-eval, /ciel-create-skill, or when skills-first-design-auditor is needed to lint a new skill. Analyzes recent sessions, runs binary evals, proposes skill patch-sets for user approval — never rewrites autonomously.
|
|
3
3
|
mode: subagent
|
|
4
|
-
model: anthropic/claude-sonnet-4-6
|
|
5
4
|
temperature: 0.2
|
|
6
5
|
tools:
|
|
7
6
|
write: false
|
|
@@ -1,7 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
description: Isolated-context researcher subagent for Ciel. Dispatch for RECHERCHE step (Standard + Critical tasks) — official docs, anti-patterns, framework philosophy, version changelog, source credibility. Also owns doc-validator-official (anti-hallucination API check). WebFetch + WebSearch enabled, no write/edit/bash.
|
|
3
3
|
mode: subagent
|
|
4
|
-
model: anthropic/claude-haiku-4-5-20251001
|
|
5
4
|
temperature: 0.2
|
|
6
5
|
tools:
|
|
7
6
|
write: false
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@neikyun/ciel",
|
|
3
|
-
"version": "6.0
|
|
3
|
+
"version": "6.1.0",
|
|
4
4
|
"description": "Ciel \u2014 Deep-reasoning pipeline for LLM-assisted development. OpenCode plugin + multi-platform CLI (OpenCode, Claude Code, more).",
|
|
5
5
|
"main": "./dist/plugin/index.js",
|
|
6
6
|
"types": "./dist/plugin/index.d.ts",
|