curx-pre-impl-routine
Pre-implementation routine for CURX tasks: read PRD/MOC, verify benchmark, update docs, implement atomically, verify proof-of-life.
Skill by gerivdb
Chalk Labs
Need help deploying curx-pre-impl-routine?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/gerivdb-curx-pre-impl-routine
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/curx-pre-impl-routine
What's inside
- Skill name
curx-pre-impl-routine- Description
- Pre-implementation routine for CURX tasks: read PRD/MOC, verify benchmark, update docs, implement atomically, verify proof-of-life. Use before any CURX implementation.
- License
- MIT
- Instructions
- 95 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# curx-pre-impl-routine > **IntentHash**: `0xCURX_PRE_IMPL_ROUTINE_20260912` > **Version**: 1.0.0 > **Domain**: CURX --- ## Objectif Documenter et executer une pre-routine systemique **avant toute implementation** CURX visant a ameliorer le benchmark score. --- ## Déclencheur Toute tache d'implementation CURX touchant : - Les metriques benchmark (`causal_accuracy`, `fragility`, `generalization`, `counterfactual_depth`) - Le score global CURX - Les engines optionnelles - Les tests d'integration --- ## Protocole obligatoire ### Étape 1 — Lire les PRD/MOC et documents de reference ``` GET C:\DevTools\CURX\PRD-CURX\*.md GET C:\DevTools\PRD-MOC\PRD-MOC-CURX-*.md GET C:\DevTools\CURX_IDEAS.md GET C:\DevTools\INTENT-MAGISTRAL-CURX-DAG.md → Verifier que le plan d'amelioration 100% est documente → Si absent : STOP — completer les PRD/MOC avant implementation ``` ### Étape 2 — Verifier l'etat actuel du benchmark ```powershell cd D:\DO\WEB\TOOLS\L0-CANON\VOLTX python scorecard.py --curx → Noter les scores actuels par niveau → Identifier les metriques < cible 100% ``` ### Étape 3 — Mettre a jour les PRD/MOC avec le plan d'amelioration - Ajouter un tableau "Plan d'Amelioration 100%" dans chaque PRD/MOC concerne - Documenter les actions par metrique - Mettre a jour les criteres d'acceptation ### Étape 4 — Implementer atomiquement - Une action = un fichier = un commit - Maximum 3 fichiers modifies entre deux commits - Ne JAMAIS attendre plus de 30 minutes sans commit ### Étape 5 — Verifier la preuve d'execution ```powershell python scorecard.py --curx → Tous les niveaux doivent afficher ✅ PASS → Score global ≥ 100/100 ``` ### Étape 6 — Mettre a jour le journal de bord - Mettre a jour `CURX_IDEAS.md` section 15 - Commit avec IntentHash horodate --- ## Checklist avant implementation - [ ] PRD/MOC lus et a jour avec plan 100% - [ ] Benchmark actuel execute et scores notes - [ ] Actions d'amelioration identifiees par metrique - [ ] Tests existants verifies (19/19 TALEX + 22/22 KIVA-CLI) - [ ] FLEX live confirme (`/health` = 200) - [ ] KG-L live confirme (`/health` = 200) --- ## Anti-patterns bloquants - Implementer sans lire les PRD/MOC d'abord - Modifier les seuils sans documenter le plan d'amelioration - Commit multiple sans preuve d'execution entre chaque - Ignorer les tests existants avant modification --- ## Référence ADR - **ADR** : ADR-2026-09-12-CURX-PRE-IMPL-ROUTINE - **IntentHash** : `0xCURX_PRE_IMPL_ROUTINE_20260912` - **Dépôt** : gerivdb/DevTools - **Statut ADR** : proposed
From gerivdb/SKILLS, MIT.
More from gerivdb/SKILLS
- 0.547score
devtools-branch-taxonomy
gerivdb
Enforce DevTools-specific branch naming: authorized prefixes and unified taxonomy format.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
branch-taxonomy-enforcer
gerivdb
Enforce repo-specific branch naming taxonomy.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
pre-push-path-audit
gerivdb
Audit paths before git push: verify local_path ↔ remote full_name consistency, detect untracked cross-repo files, and ensure no forbidden paths are pushed.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
git-remote-safety-guard
gerivdb
Verify git remote safety before push: correct remote URL, correct branch, no force push without verification.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
git-workflow-guardian
gerivdb
Enforce correct git workflow: never push main directly, use feature branches + PR, respect repo-specific branch taxonomy and ALFRED/BRGS guards.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
session-workflow-closer
gerivdb
End-of-session workflow: verify no orphan branches, ensure all feat branches have PRs or are merged, clean up local branches, and confirm no direct main push…
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
artifact-quality-checker
gerivdb
Verifie la qualite des artefacts de gouvernance (PRD, ADR, EPIC, INTENT) selon les probes P-101..P-109.
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
codebase-intelligence
gerivdb
Orchestration de ecosystem-brain + codebase-memory-mcp pour des réponses combinées sur l'écosystème GeriCode.
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
mox-validator
gerivdb
MOX — Gardien de coherence cross-repo.
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
n243-graph-builder
gerivdb
Construit le graphe souverain cross-repo N243.
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
n243-query-engine
gerivdb
Moteur de requête N243.
SkillBFreshness: B★ 1Claude Code +9 - 0.536score
pr-review-resolver
gerivdb
Review + resolve PR via diffscope + auto-merge gate + cleanup across all ecosystem repos
SkillBFreshness: B★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 8
- Last commit (this folder)
- Sep 12, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/gerivdb/SKILLS
- Website
- github.com/gerivdb/SKILLS
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from gerivdb
- 0.547score
devtools-branch-taxonomy
gerivdb
Enforce DevTools-specific branch naming: authorized prefixes and unified taxonomy format.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
pre-push-path-audit
gerivdb
Audit paths before git push: verify local_path ↔ remote full_name consistency, detect untracked cross-repo files, and ensure no forbidden paths are pushed.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
branch-taxonomy-enforcer
gerivdb
Enforce repo-specific branch naming taxonomy.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
git-remote-safety-guard
gerivdb
Verify git remote safety before push: correct remote URL, correct branch, no force push without verification.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
git-workflow-guardian
gerivdb
Enforce correct git workflow: never push main directly, use feature branches + PR, respect repo-specific branch taxonomy and ALFRED/BRGS guards.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
session-workflow-closer
gerivdb
End-of-session workflow: verify no orphan branches, ensure all feat branches have PRs or are merged, clean up local branches, and confirm no direct main push…
SkillBFreshness: B★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get curx-pre-impl-routine running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
curx-pre-impl-routine skill. Agent Mart. https://agentmart.dev/skills/gerivdb-curx-pre-impl-routine (last updated 2026-10-24)