Something new from YAC Yazılım: — an AI CEO that runs your company. Coming soon. And here's the part we're proud of — not one of these was built by someone who writes code: → — an AI CEO for your business → — turn a drawing into a 3D scaffold design + bill of materials → — a B2B scaffolding & formwork marketplace → — an AI toolkit for small businesses All built from İstanbul, with @AnthropicAI's Claude Code. You bring the body — we provide the brain. #ClaudeCode# #Anthropic# #AI#

🧠 Claude Code agents start every session with amnesia. This post solves that without adding a dedicated service, using the Elasticsearch you already run. TL;DR: A CLI called bridge and three hooks automatically store an agent's decisions, context, and tasks in Elasticsearch, then recall them across sessions and devices via hybrid search plus temporal decay. Title: Persistent memory for agents: Claude Code on Elasticsearch URL: Points ・🗂 Seven indices: memory/messages/tasks/sessions/status/entities/entity-history store memory by dimension ・🪝 Three hooks automate it: SessionStart syncs, markdown writes get indexed, Stop logs the session end (no explicit calls) ・🔎 Hybrid recall: BM25 fused with semantic_text dense vectors via RRF, catching both exact task IDs and conceptual matches ・⏳ Temporal decay: a default 45-day DECAY ranks recent memories higher (needs ES 9.3+ or Serverless) ・🕸 Knowledge graph: extracts blocked_by and friends from markdown frontmatter, surfacing blockers at depth-2 traversal ・📡 Offline resilience: writes queue locally as JSON and flush via bulk API once connectivity returns ・💻 Cross-device: gen-handoff produces a handoff JSON so another machine restores context with no git pull If you already run Elasticsearch, the pragmatic appeal is no new service and reusing your existing monitoring and backups. Setup is three commands: git clone and #Elasticsearch# #ClaudeCode#

Your Claude Code work becomes an always-up-to-date live shared page — Artifacts eliminate the time spent explaining status 📊 Title: Claude Code now supports artifacts URL: 📊 Overview Artifacts turn a Claude Code work session into a live, interactive web page. They pull in the full session context — codebase, connectors, and conversation — and output a visual, shareable result. ❓ Challenges Solved Dev teams spend more time communicating status than building. ・Manually compiling findings, explaining progress verbally, recreating visualizations ・This overhead drags on development Artifacts automate it and keep the page synchronized as work progresses. 💡 Features & How It Works ・Built on full session context: code, monitoring data, and reasoning combined with no external infrastructure ・Live pages that update in place: teammates see refreshes immediately on publish ・Version history for restoring previous states, plus a gallery to manage them ・Private by default, shareable with org members, with admin-controlled access The flow: request an artifact → Claude Code builds and publishes it → share the link → it auto-republishes to the same URL as work continues. 🌍 Use Cases License audits, data-flow mapping, security findings linked to code lines, cloud cost tracking, PR walkthroughs, architecture diagrams, incident postmortems, and team shipping summaries. #ClaudeCode# #AIAgents#

# Claude Code Features and Practical Usage 🔒 Control exactly how autonomously Claude operates. Read-only for untrusted investigation, auto-approve in trusted environments: safety and speed at once. 📌 Title and Feature URL Title: Permissions / Sandbox URL: 📝 Overview Claude Code provides a fine-grained permission system so you can specify exactly what the agent may and may not do. Combine allow/ask/deny rules with permission modes, check the config into version control, and distribute the same guardrails to your whole team. A sandbox adds OS-level isolation on top. 🔧 How It Works - A tiered model: read-only tools (file reads, Grep) need no approval; Bash commands and file edits require approval. - Rules evaluate in order: deny, then ask, then allow. The first match wins. A deny at any scope takes precedence and cannot be overridden by an allow. - Permission modes include default, acceptEdits, plan, auto, dontAsk, and bypassPermissions. plan only reads and runs read-only commands; bypassPermissions is nearly all-permissive and risky. - Rules are enforced by Claude Code, not the model. Prompt or CLAUDE.md instructions shape what Claude tries, not what it is allowed to do. 🛠 Practical Usage - During a session, use "/permissions" to view and edit all rules and the settings.json they come from. - Rule syntax is "Tool" or "Tool(specifier)". Examples: Bash(npm run build), Read(./.env), WebFetch(domain: Bash supports glob "*". - Configure in settings.json under permissions.allow / deny. Example: allow "Bash(npm run *)", deny "Bash(git push *)" and "Read(./.env)". - For read-only investigation, use plan mode, or deny "Edit" and "Write" and narrow Bash. Control subagents with rules like Agent(Explore). - Grant extra directory access via "--add-dir" or "/add-dir", and persist it with additionalDirectories. 🎯 Use Cases - Investigate an untrusted repo in plan mode to understand structure without modifying source. - Define safe allow/deny in a shared .claude/settings.json so everyone gets the same guardrails. - Inside a container or VM, use bypassPermissions to skip prompts and speed up CI or bulk work. - Restrict where Bash can reach using the sandbox's ⚠️ Caveats - bypassPermissions skips prompts even for writes to .git and .claude, so only use it in isolated environments (rm -rf / still stops as a circuit breaker). - Argument-constraining curl permission patterns are fragile. It is more reliable to deny Bash network tools and allow specific WebFetch(domain:...) entries. - Read/Edit deny rules apply to built-in file tools and recognized Bash commands only, not to files a Python or Node script opens directly. Use the sandbox to bind all processes. - Managed settings cannot be overridden by any other scope; enforce organizational policy there. #ClaudeCode# #DevTools#

# Claude Code Features and Practical Usage 🚀 Just mention "@/claude" in a PR or issue and watch fixes, features, and reviews happen automatically. Claude Code becomes an AI teammate that works inside your CI. 📌 Title and Feature URL Title: GitHub Actions URL: 📝 Overview Claude Code GitHub Actions integrates Claude Code into your GitHub workflow. Mention "@/claude" on any PR or issue and Claude analyzes your code, creates pull requests, implements features, or fixes bugs. Built on the Claude Agent SDK, it is ideal for automating routine engineering tasks. 🔧 How It Works - It auto-detects whether to run in interactive mode (responding to @/claude mentions) or automation mode (running immediately with a prompt) based on your config. - It respects the repository's CLAUDE.md, following your project standards and existing code patterns. - Code runs on GitHub-hosted runners and uses Sonnet by default (Opus 4.8 can be selected). - Beyond the direct Claude API, it supports Amazon Bedrock and Google Vertex AI. 🛠 Practical Usage - The easiest setup is running "/install-github-app" inside claude in your terminal; it guides you through the GitHub app and secrets (you need repo admin rights). - For manual setup: install the Claude GitHub app ( add ANTHROPIC_API_KEY as a repository secret, and copy examples/claude.yml into .github/workflows/. - Use the action "anthropics/claude-code-action@v1". Pass instructions via "prompt" and CLI arguments via "claude_args". - claude_args examples: --max-turns 5 / --model claude-sonnet-4-6 / --mcp-config /path/to/config.json - Comment examples: "@/claude implement this feature based on the issue description", "@/claude fix the TypeError in the user dashboard component". 🎯 Use Cases - Mention on an issue to auto-create a PR that turns requirements into working code. - Request a standard review on a PR: "@/claude review this for security issues". - Use a schedule (cron) trigger to auto-generate a daily summary of yesterday's commits and open issues. - Wire in the code-review plugin to run its skill automatically on every PR update. ⚠️ Caveats - Never commit API keys directly to the repo; always reference GitHub Secrets (secrets.ANTHROPIC_API_KEY). - Costs accrue both as GitHub Actions minutes and API tokens. Use --max-turns and timeouts to prevent runaway jobs. - If Claude doesn't respond, confirm the comment uses "@/claude" (not "/claude") and that the app and secrets are configured. - v1.0 has breaking changes from beta: remove mode, change direct_prompt to prompt, and move CLI options into claude_args. #ClaudeCode# #CICD#

# Claude Code Features and Practical Usage 🔌 Still copy-pasting schemas and logs into chat? Connect Sentry, PostgreSQL, or GitHub via MCP and Claude reads and writes those systems directly. "Investigate the last 30 days of this table" just works. 📌 Title and Feature URL Title: MCP URL: 📝 Overview MCP (Model Context Protocol) is an open standard for AI tool integration that connects Claude Code to hundreds of external tools and data sources. MCP servers give Claude Code access to tools, databases, and APIs, so it can read and operate those systems directly instead of working from what you paste. 🔧 How It Works - Multiple transports exist: HTTP (recommended) for remote connections, WebSocket for bidirectional push, stdio for local processes, and the deprecated SSE. - Three scopes are available: local (default, private to you, stored in ~/.claude.json), project (shared via .mcp.json), and user (all projects, private to you). - Many cloud servers require authentication and support OAuth 2.0; on a 401/403 response you can complete the flow from /mcp. - Tool search is on by default, lazily loading MCP tools on demand to keep context usage low. 🛠 Practical Usage - Add with claude mcp add --transport http , or for stdio claude mcp add [options] -- [args...]. Options go before the server name. - Manage with claude mcp list / claude mcp get / claude mcp remove , and use /mcp inside a session to check status and authenticate. - Set scope with flags like --scope project; for project sharing, the .mcp.json file is version-controlled. - Reference MCP resources with @/server:protocol://resource/path, e.g. analyze @/github:issue://123. - Run MCP prompts as commands in the form /mcp__servername__promptname. 🎯 Use Cases - Connect Sentry and debug production issues: "What are the most common errors in the last 24 hours?" - Connect PostgreSQL and query without pasting the schema: "Find customers who haven't purchased in the last 90 days." - Connect GitHub's remote MCP and ask "Review PR #456# and suggest improvements." - Commit .mcp.json so the whole team shares the same set of MCP tools. ⚠️ Caveats - Confirm you trust each server before connecting; servers that fetch external content can be exposed to prompt-injection risks. - Project-scoped servers require approval before use for safety (reset choices with claude mcp reset-project-choices). - A warning appears when tool output exceeds 10,000 tokens; raise the cap with MAX_MCP_OUTPUT_TOKENS (default max is 25,000). - SSE is deprecated; use HTTP where possible. #ClaudeCode# #MCP#