Use Fledge from an AI agent

Fledge ships an Agent Skill that teaches an LLM agent the inspect → fix → check → prepare loop, how to read the compact and JSON reports, what each exit code means, and which safety rules apply. The skill only describes the installed fledge command; it adds no network access or permissions.

Install the skill

From a source checkout or extracted source distribution, copy the skill folder into the agent’s skill directory. For Claude Code:

# For all your projects:
mkdir -p ~/.claude/skills && cp -R skills/fledge ~/.claude/skills/
# Or only for one paper repository:
mkdir -p /path/to/paper/.claude/skills && cp -R skills/fledge /path/to/paper/.claude/skills/

Other agents that support the Agent Skills format can load skills/fledge/SKILL.md the same way. The fledge executable must be on the agent’s PATH; see installation for the launcher location.

What the agent runs

The skill asks the agent to prefer one-line-per-finding output:

fledge check /path/to/paper --main main.tex --output-format compact --quiet
fledge rule TEX005 --json

Use --json when the agent needs complete evidence, diffs or tool versions. To apply venue-oriented checks, the skill uses --preset NAME and narrows runs with --select and --ignore. The skill tells the agent not to edit scientific content to silence heuristics, not to enable online checks without asking, and to report a submission as ready only after prepare released submission.zip. A test keeps every command and option named in the skill in sync with the CLI.