Agently技能

Install

Choose the target agent first. The recommended path is to install one bundle into one agent-specific skill directory, for example Codex:

export AGENT=codex

Use AGENT=claude, AGENT=cursor, or another supported agent when that is your actual target.

app Default bundle for building new Agently applications:

for skill in \
  agently-playbook \
  agently-request \
  agently-runtime \
  agently-dynamic-task \
  agently-triggerflow
do
  npx skills add AgentEra/Agently-Skills --agent "$AGENT" --skill "$skill" -y
done

migration Bundle for moving existing LangChain, LangGraph, LlamaIndex, CrewAI, or similar systems into Agently. Install the app bundle first, then add the migration skill:

npx skills add AgentEra/Agently-Skills --agent "$AGENT" --skill agently-migration -y

Install only the router when you want the smallest possible starting point:

npx skills add AgentEra/Agently-Skills --agent "$AGENT" --skill agently-playbook -y

Inspect the default public catalog:

npx skills add . --list

The default listing and standard install path expose only the current 6-skill catalog.

Optional Companion Package

Agently keeps agently-devtools as an optional developer-tooling companion package.

pip install agently-devtools
agently-devtools init my_project

- Install: pip install -U agently agently-devtools - Compatibility line: use the Agently compatibility registry for the active release or development target - Public entrypoints: ObservationBridge, EvaluationBridge, EvaluationRunner, and create_local_observation_app - Recommended startup: agently-devtools start

Use this package when an Agently app needs local runtime observation, evaluations, logs, or playground support during development and debugging. The skills package treats this as optional observability tooling, not as a required source-repo dependency.

Recommended observation wiring binds at bridge creation time and selects scope with watch(...):

from agently import Agently
from agently_devtools import ObservationBridge

bridge = ObservationBridge(Agently, app_id="your_app_id")
bridge.watch(agent)