Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
/plugin marketplace add fledgeling-co/fledgeling-pluginsnpx agentmods add plugins/fledgeling-co/fledgeling-plugins/flagshipgit clone --depth 1 https://github.com/fledgeling-co/fledgeling-pluginsGrade A, and why
flagship scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "flagship",
"version": "1.6.5",
"description": "Conduct many concurrent Claude Code sessions as one portfolio — the layer above ship-armada, for when the work is already spread across a dozen live sessions rather than waiting in one repo's backlog. Builds a roster of every peer session, learns what each owns, and coordinates them over SendMessage: hands out a capped heavy-work token derived from harbourmaster's measured berths rather than an invented number, batches their accumulated decisions through the perch R1–R5 router so only the operator's own axes reach the operator, routes judgement to defer's lanes after probing which are actually reachable, and propagates each session's findings to the others — which is the real return, because nine sessions independently found one tool's bug in a single evening and none knew about the others. Starts new work as a workflow, a subagent, or a fresh Ghostty tab using recover-claude-code's hardened mechanism, so a session can carry its own fleet. Holds the map and no authority: a conducting session cannot command a peer, cannot close a peer's channel to its user, and cannot relay an authorisation — all three were refused by ten sessions in one night and all three refusals were right.",
"category": "development"
}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 7 lines scan A 4a2d0f65c901
flagship is a plugin published in the GitHub repository fledgeling-co/fledgeling-plugins (2 stars, last pushed yesterday), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other plugins, from other repositories
jira-assistant-skills
Plugin marketplace listing 1 plugin: jira-assistant-skills.
beans-prime
Configure Claude Code to track work using Beans.
rook-claude
Claude Code plugins for rook (github.com/rook/) maintainers: maintainer-grade code review, backlog triage, systemic-PR campaigns, and rook house conventions.
rook-maintainer
Maintainer-grade skills for rook (github.com/rook/): code review, backlog triage, systemic-PR campaigns, and the rook house conventions they enforce.
ai-skills
Plugin marketplace listing 8 plugins: security-hooks, pr-review-triage, dev-skills, nix-skills, git-workflow.
workbench
Ticket-to-PR development workflow: triage, plan with verified citations, implement inside the plan, review, commit, PR. Pluggable issue trackers (Jira, Azure DevOps, GitHub Issues) or a backlog in the repo, so it works with no tracker at all.