Autoresearch Wrapper Flow

Use the run.sh launcher in this skill's folder (the folder containing this SKILL.md). This block finds it in the standard Claude Code and Codex install locations:

AUTORESEARCH_RUNNER=""
if [ -n "${CLAUDE_SKILL_DIR:-}" ]; then
  AUTORESEARCH_RUNNER="${CLAUDE_SKILL_DIR}/run.sh"
elif [ -n "${CLAUDE_PLUGIN_ROOT:-}" ]; then
  AUTORESEARCH_RUNNER="${CLAUDE_PLUGIN_ROOT}/skills/autoresearch-wrapper-flow/run.sh"
else
  for dir in skills/autoresearch-wrapper-flow .agents/skills/autoresearch-wrapper-flow .claude/skills/autoresearch-wrapper-flow "$HOME/.agents/skills/autoresearch-wrapper-flow" "$HOME/.claude/skills/autoresearch-wrapper-flow" "$HOME/.codex/skills/autoresearch-wrapper-flow"; do
    if [ -f "$dir/run.sh" ]; then AUTORESEARCH_RUNNER="$(readlink -f "$dir/run.sh")"; break; fi
  done
fi
[ -n "$AUTORESEARCH_RUNNER" ] || { echo "Could not resolve the autoresearch-wrapper-flow launcher" >&2; exit 1; }

Use:

bash "$AUTORESEARCH_RUNNER" flow

When this skill is invoked:

1. Show the recorded metric flow for the active run by default.
2. Summarize:

• metric name
• goal
• chronological sequence
• best-so-far sequence
• latest metric
• best metric

3. Include the ASCII plot when presenting the result.

4. If the user asks for raw structured data, use:

bash "$AUTORESEARCH_RUNNER" flow --json

Example

• /autoresearch-wrapper-flow