hep-call

SkillAI & models

Once this skill is added, your AI can prepare Agentlas Hub or Cloud agents that you name explicitly. That means you can point to a specific agent by name and have it prepared for use. The skill comes from the agentlas-os project on GitHub.

Available today. Use it from your connected AI after setup.

After adding the skill, tell your AI the name of the Agentlas Hub or Cloud agent you want prepared. It will take care of the preparation from there.

Then ask your AI: use the hep-call skill

What your AI can do with it

  • Prepare a named Agentlas Hub agent
  • Prepare a named Agentlas Cloud agent
  • Work with both Agentlas Hub and Agentlas Cloud agents
  • Target a specific agent by its exact name

What this skill tells your AI

The instructions your AI receives, as published by agentlas-ai/agentlas-os in kimi/skills/hep-call/SKILL.md and read by ahel’s review.

Update fallback: 자동 업데이트가 안 되면 hephaestus update를 한 번 실행하세요. 업데이트하지 않아도 현재 버전 명령은 그대로 동작합니다.

Hephaestus Call

Raw arguments: everything the user typed after /skill:hep-call.

Codex plugins cannot register slash commands, so this custom prompt is the explicit entrypoint: /prompts:hep-call.

  1. Resolve the runner and prepare the named agents:
RUNNER=""
for c in "$HOME/.agentlas/runtime/current/bin/hephaestus" ./bin/hephaestus; do
  [ -x "$c" ] && RUNNER="$c" && break
done
[ -n "$RUNNER" ] || { echo "Hephaestus runtime not found. Run the installer first." >&2; exit 1; }
if [ "${HEPHAESTUS_AUTH_AUTOPOPUP:-1}" != "0" ]; then
  "$RUNNER" auth ensure --timeout 180 >/dev/null 2>&1 || true
fi
RAW="$ARGUMENTS"
if printf '%s' "$RAW" | grep -q '{'; then
  AGENTS="${RAW%%\{*}"
  CONTEXT="${RAW#*\{}"
  CONTEXT="${CONTEXT%\}}"
else
  AGENTS="${RAW%% *}"
  CONTEXT="${RAW#* }"
fi
AGENTS="$(printf '%s' "$AGENTS" | sed 's/[[:space:]]*$//')"
CONTEXT="$(printf '%s' "$CONTEXT" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
"$RUNNER" call "$AGENTS" "$CONTEXT" --runtime codex
  1. For each prepared agent, follow its returned output.entry_excerpt and output.grounding.directive. The Hub returns BYOM instructions; Codex executes with the current model and permission model.

  2. Report failures separately and include the top-level receipt_id plus every prepared agent execution_id.

Signals

GitHub stars
1k
Forks
103
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
hep-call
Source
github.com/agentlas-ai/agentlas-os