Armory
Source

Browse

Search and filter by type across the catalog

262 results in Memory, Evals, Infrastructure, Hooks · page 7 of 11

HooksPreview

discord-notifications

Send Discord notifications when Claude Code finishes working. Requires DISCORD_WEBHOOK_URL environment variable. Get webhook URL from Discord Server Settings -> Integrations -> Webhooks.

UnrankedNo signals yet
Harnessclaude
automationhooks
Details
HooksPreview

edit-audit-log

Log all file edits to a project-local audit file with timestamps. Records every Edit tool usage to .claude/edit-log.txt for tracking changes across sessions. Requires jq.

UnrankedNo signals yet
Harnessclaude
development-toolshooks
Details
HooksPreview

env-file-protection

Prevent writing to .env files using the if condition for lightweight filtering. Blocks any Write tool call targeting .env* files, protecting secrets from accidental overwrites. Uses the if field to avoid spawning a process unless the file pattern matches.

UnrankedNo signals yet
Harnessclaude
securityhooks
Details
EvalsPreview

evals-cookbooks

OpenAI Cookbook eval recipes: task-specific templates for summarization, QA, and classification evaluation using the Evals framework.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
evalscookbooktemplatesopenai
No one-command install · SourceDetails
InfrastructurePreview

fermyon-spin

WebAssembly serverless framework for building and deploying fast, portable agent microservices.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructurewasm-serverless
No one-command install · SourceDetails
HooksPreview

file-backup

Automatically backup files before editing. Creates timestamped backups in a .backups directory when files are modified.

UnrankedNo signals yet
Harnessclaude
development-toolshooks
Details
HooksPreview

file-protection

Protect critical files from accidental modification. Prevents editing of important system files, configuration files, and production code.

UnrankedNo signals yet
Harnessclaude
securityhooks
Details
InfrastructurePreview

firecracker

AWS-open-sourced microVM technology powering secure, fast (<125ms) isolation for serverless and container workloads.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructuremicrovm
No one-command install · SourceDetails
InfrastructurePreview

fly-io

Deploy full-stack apps and long-running agent processes globally via Firecracker microVMs close to users.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructuredeploy
No one-command install · SourceDetails
HooksPreview

force-push-blocker

Block git force push commands using the if condition for efficient filtering. Prevents accidental force pushes that can overwrite remote history. Covers --force, --force-with-lease, and the -f shorthand.

UnrankedNo signals yet
Harnessclaude
securityhooks
Details
HooksPreview

format-javascript-files

Automatically format JavaScript/TypeScript files after any Edit operation using prettier. This hook runs 'npx prettier --write' on any .js, .ts, .jsx, or .tsx file that Claude modifies, ensuring consistent code formatting. Uses npx so prettier doesn't need to be globally installed. Includes error suppression so it won't fail if prettier is not available.

UnrankedNo signals yet
Harnessclaude
post-toolhooks
Details
HooksPreview

format-python-files

Automatically format Python files after any Edit operation using black formatter. This hook runs 'black' on any .py file that Claude modifies, ensuring consistent Python code formatting. Requires black to be installed ('pip install black'). The command includes error suppression (2>/dev/null || true) so it won't fail if black is not installed.

UnrankedNo signals yet
Harnessclaude
post-toolhooks
Details
EvalsPreview

gaia-benchmark

GAIA: benchmark of 466 real-world questions requiring multi-step reasoning, web browsing, and tool use for general AI assistants.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
evalsbenchmarkagentstool-use
No one-command install · SourceDetails
HooksPreview

git-add-changes

Automatically stage changes in git after file modifications for easier commit workflow. This hook runs 'git add' on any file that Claude edits or writes, automatically staging changes for the next commit. Includes error suppression so it won't fail in non-git repositories. Helps streamline the development workflow by preparing files for commit.

UnrankedNo signals yet
Harnessclaude
post-toolhooks
Details
InfrastructurePreview

gvisor

Google-open-sourced application kernel providing a sandboxed container runtime for untrusted agent code.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructuresandbox
No one-command install · SourceDetails
EvalsPreview

honeyhive

LLM evaluation and experimentation platform with session tracing, dataset management, and metric-based run comparison.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
evalsexperiment-trackingtracingdataset
No one-command install · SourceDetails
InfrastructurePreview

kata-containers

Lightweight VMs that combine container speed with VM-level security isolation for agent runtime workloads.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructuremicrovm
No one-command install · SourceDetails
InfrastructurePreview

kubernetes

Industry-standard container orchestration system for deploying, scaling, and managing agent workload fleets.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
infrastructureorchestration
No one-command install · SourceDetails
HooksPreview

langsmith-tracing

Automatically send Claude Code conversation traces to LangSmith for monitoring and analysis. Prerequisites: jq (brew install jq on macOS or sudo apt-get install jq on Linux), curl and uuidgen (usually pre-installed), LangSmith account and API key. Configuration: Install langsmith-tracing setting (npx claude-code-templates@latest --setting telemetry/langsmith-tracing) or manually add to .claude/settings.local.json the following environment variables: TRACE_TO_LANGSMITH=true, CC_LANGSMITH_API_KEY=lsv2_pt_..., CC_LANGSMITH_PROJECT=project-name, CC_LANGSMITH_DEBUG=true (optional). How it works: Runs in background on Stop event after each Claude response, reads conversation transcript, converts to LangSmith format, sends to LangSmith API, groups by thread_id for session continuity. Debugging: Check logs at ~/.claude/state/hook.log. Privacy note: System prompts not included in traces.

UnrankedNo signals yet
Harnessclaude
monitoringhooks
Details
HooksPreview

langsmith-tracing-2

#

UnrankedNo signals yet
Harnessclaude
monitoringhooks
Details
InfrastructurePreview

lightpanda

Ultra-fast headless browser written in Zig specifically for AI and automation workloads. It runs JavaScript natively, uses 9x less memory than Chrome headless, and targets sub-100ms page execution.

UnrankedNo signals yet
Harnessclaudecursorcodexopencodegemini
browserlightpanda
No one-command install · SourceDetails
HooksPreview

lint-on-save

Automatically run linting tools after file modifications. Supports ESLint for JavaScript/TypeScript, Pylint for Python, and RuboCop for Ruby.

UnrankedNo signals yet
Harnessclaude
development-toolshooks
Details
HooksPreview

memory-persistence

Lifecycle hook bundle (SessionStart, PreCompact, observation, SessionEnd) that loads prior context and persists session state across compaction.

UnrankedNo signals yet
Harnessclaudecodexcursorgeminiopencode
hooksmemorylifecycle
Details
HooksPreview

memory-persistence-bundle

Hook bundle: Reference lifecycle hook definitions for ECC memory persistence. The production hook graph is hooks/hooks.json.. Use to persist and reload agent session state across compaction.

UnrankedNo signals yet
Harnessclaudecodexcursorgeminiopencode
hookmemory-persistence
Details
Browse · Armory