Kiro Crew is a persistent development workspace where agents continue multi-step software work across sessions, schedules, and connected interfaces. Developers use it locally or remotely through a desktop app, web dashboard, CLI, Slack, or Discord, with unattended tasks and recurring jobs. The catalogue contains skills and instructions for working with this workspace.
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.
npx skills add kirodotdev/KiroCrew --skill session-summariesgit clone --depth 1 https://github.com/kirodotdev/KiroCrewWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kirodotdev/kirocrew/session-summaries)<a href="https://agentmods.dev/skills/kirodotdev/kirocrew/session-summaries"><img src="https://agentmods.dev/badge/skills/kirodotdev/kirocrew/session-summaries/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kirodotdev/kirocrew/session-summaries"><img src="https://agentmods.dev/badge/skills/kirodotdev/kirocrew/session-summaries.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00053 | $0.01339 |
| Opus 5.5 | $0.00021 | $0.00536 |
| Sonnet 5.5 | $0.00011 | $0.00268 |
| Haiku 4.5 | $0.00005 | $0.00134 |
Grade A, and why
session-summaries 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 11d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session summaries
The session summary panel lives in the chat right side panel. When enabled, it distills a session into the few things that make returning to it cheap: what the person was trying to do, what is actually true now, and what is still open. It is generated by a separate background pass after a turn completes — you do not write it, and reading this skill does not trigger it.
Use this skill to explain the panel, diagnose why it looks empty or wrong, and help the person get more out of it. It changes how you talk about the feature, never how the summary is produced.
Turning it on
It is off by default because summarizing spends tokens on a turn the person did not ask to pay for. The single opt-in is Settings → Chat → Sessions → Session summaries. Point them there rather than editing config by hand.
- An unchanged session is served from cache for free — the cost is only on turns that actually change the session.
- The cadence and caps (minimum user turns before the first summary, how often it
rebuilds, how many intents it keeps) are power-user knobs in
config.jsonundersession_summary.*, not in the UI. Mention them only if asked.
What the panel is built from
The summary is organized by intent — a goal the person pursued, which can span many turns. A session that pivoted has several intents; a topic label is not an intent. Each intent carries:
- status — about the work:
active,completed, orabandoned. - verified — about the result, independent of status: was the outcome ever
confirmed? Shipped-but-never-run, diagnosed-but-never-fixed, and merged-but-never-
seen are all
completedyet unverified. This is the panel's most useful signal; it is exactly the work a person forgets. It is three-valued:trueconfirmed,falseshipped-but-unconfirmed, andnullwhen confirmation does not apply — so an absent verification is not the same as a failed one. The panel never shows the two fields separately: it shows one chip, Done, Needs you (completed butfalse), Active or Dropped. - progress — a short runbook of what is true now and known to work, not a turn-by-turn history. Shorter is better here.
- next steps — inferred open actions, each with why it matters and what to expect. An intent that needs nothing has none.
- project notes — a handful of session-scoped operational facts (a required flag, a step that must follow a change, a name you were corrected on). Durable cross-session preferences belong in lessons, not here.
- turn ranges — each intent shows which turns it spans, how many times the session returned to it, and the turn it pivoted from, so a session that ping-ponged between two goals reads as such.
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.
- 11d ago Changed · +7 lines 7d3252f6564f
- 1mo ago Changed · +19 lines 426c7b9621de
- 1mo ago First seen · 75 lines · 53 tokens per session scan A 469184987459
session-summaries is a skill published in the GitHub repository kirodotdev/KiroCrew (4,481 stars, last pushed today), licensed Apache-2.0. It adds 53 tokens to every session and 1,339 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-30.
Other skills, from other repositories
persistent-notes
Save notes locally to /mnt/workspace/notes.json file. Use when user wants to "save a note" or "remember something".
search-shared-sessions
Search an Agent Relay workspace's shared coding-agent sessions for prior work, decisions, fixes, or the context behind a task or pull request. Use when team history could answer the user's question or prevent repeating an investigation.
session-sync
Syncs Claude Code project data — memories, session transcripts, settings, agent definitions, remember-logs, and handoff documents — between the same project on a remote workstation and the local machine over SSH, so local work can resume exactly where a remote session left off. Use when starting on a new machine…
list-learned-actions
Explicit Codex workflow: List persisted reusable actions, UI skeletons, and legacy feedback memories before composing device primitives.
hindsight-local
Store user preferences, learnings from tasks, and procedure outcomes. Use to remember what works and recall context before new tasks. (user).
memory-context
Load relevant long-term memory from Hindsight before starting work. Use at the start of every task or request, before planning or answering, and whenever the user mentions people, projects, decisions, preferences or anything done before.