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 zookanalytics/gc-toolkit --skill converse-recheck-premisegit clone --depth 1 https://github.com/zookanalytics/gc-toolkitWrote 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/zookanalytics/gc-toolkit/converse-recheck-premise)<a href="https://agentmods.dev/skills/zookanalytics/gc-toolkit/converse-recheck-premise"><img src="https://agentmods.dev/badge/skills/zookanalytics/gc-toolkit/converse-recheck-premise/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/zookanalytics/gc-toolkit/converse-recheck-premise"><img src="https://agentmods.dev/badge/skills/zookanalytics/gc-toolkit/converse-recheck-premise.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00022 | $0.00710 |
| Opus 5.5 | $0.00009 | $0.00284 |
| Sonnet 5.5 | $0.00004 | $0.00142 |
| Haiku 4.5 | $0.00002 | $0.00071 |
Grade A, and why
converse-recheck-premise 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 10d 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.
What it actually says
Step 2 — Re-check the premise
The condition that justified filing a visit routinely dies before anyone claims it. Test the VISIT's own premise against live state before you prep, and before the rename: a visit that closes here should not have moved the operator's session title.
$CONV, $VISIT, and $SUBJECT are resolved in step 1's claim block.
Re-read the visit body. Its stated conditions ARE the premise, often
bulleted literally — "no triage.hold and no gc.takeaway on the
root", "its frontier is [...] UNASSIGNED". Check each one still
holds, on the subject and on whatever bead the premise is about:
gc bd show "$SUBJECT" --json | jq -r '.[0].metadata
| "hold=\(.["triage.hold"] // "") takeaway=\(.["gc.takeaway"] // "")"'
NON-EMPTY is the test for triage.hold — an EMPTY stamp is a CLEARED
hold. A gc.takeaway dates the last sitting rather than naming a live
wait: read what it says, then check whether that wait is still open.
Two readings end the visit here, with nothing posted:
-
moot — the premise no longer holds. The frontier was routed, the bead was closed, another visit already settled it.
-
benign — the premise holds but needs no human: the wait is already named by a non-empty
triage.holdor an open demand bead, or the condition is a known acceptable state. An open PR awaiting the operator's review is the canonical case — their own review queue, and handing it back is the bug this step prevents.A takeaway is never a benign wait on its own, because nothing clears it. Re-check the ids in the body (
bd show, andgc bd list --parent "$SUBJECT" --all); it is moot only if something is open again.
Close it out with converse-close-out.sh, which appends the reading to
the subject's notes, stamps gc.outcome=<moot|benign> on the visit,
reads it back, and closes the visit — no takeaway, nothing posted:
VISIT="$VISIT" SUBJECT="$SUBJECT" \
"$CONV/converse-close-out.sh" <moot|benign> "<the premise, and what is true instead>"
Then go to step 8 and claim again. Post nothing — no framing, no sign-off, not even "this turned out to be fine". Deliberately no takeaway stamp either: it is the subject's headline of what it NEEDS, and one for a visit that needs nobody spends the attention this exit saves.
The exit is gated on being named: if you cannot point at the stamp or state that makes this benign, you hold the sitting. Uncertain is not benign.
When the premise still holds and needs a human, this visit earns a sitting:
prep it (steps 3–4, the converse-prep skill).
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.
- 10d ago Changed · -1 lines 8262065fc238
- 26d ago First seen · 62 lines · 22 tokens per session scan A de3ecdf5dbfe
converse-recheck-premise is a skill published in the GitHub repository zookanalytics/gc-toolkit (5 stars, last pushed today), licensed MIT. It adds 22 tokens to every session and 710 once invoked, about $0.0001 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 2026-09-22.
Other skills, from other repositories
dispatch
Use when a task file exists in .hyperflow/tasks/ and workers need dispatching. Fans out parallel workers under per-batch Reviewers, runs a final integration review, and commits per sub-task. Endpoint of the auto-chain — no auto-deploy. Trigger with /hyperflow:dispatch, "run the plan", "execute the task", "build it"…
commonly
You are a member of a Commonly workspace — a shared space where humans and AI agents from any origin collaborate in pods (chat rooms with memory). Use this whenever you are connected to Commonly via the commonly MCP tools: to read what's happening, post, remember things across sessions, react, DM other agents, and…
sdlc-accelerate
End-to-end SDLC ramp-up from idea to construction-ready with automated phase transitions.
project-awareness
Detect project type, AIWG framework state, team configuration, and active work to summarize status and recommend next actions.
artifact-metadata
Manage artifact metadata, versioning, ownership, and review history across the SDLC lifecycle.
mission-control
Orchestrate multi-loop background operations via the Mission Control dashboard — start sessions, dispatch missions, monitor, and stop.