slack-channel-summarizer
Skill142 starsChanged 3 months agoApache-2.0View on GitHub
Read and summarize Slack channel history from inside the NemoClaw sandbox.
In this repository62 other files it gives its agents.
- examples/recipes/partners/hpe/retail-assistant/docker-compose/scripts/identity/AGENTS.mdAGENTS.md
- examples/recipes/partners/hpe/retail-assistant/helm/files/identity/AGENTS.mdAGENTS.md
- examples/recipes/partners/tavily/watchtower/prompts/AGENTS.mdAGENTS.md
- nemoclaw-community-contributor-examplesSkill
- nemoclaw-community-maintainer-policiesSkill
- coach-nemoclaw-hermesSkill
- coordinate-nemoclaw-blenderSkill
- blender-host-sandbox-boundarySkill
- blender-python-api-verificationSkill
- ovphysx-host-runtime-boundarySkill
- axe-a11ySkill
- deep-researchSkill
- cross-source-gap-analysisSkill
- github-readonly-liveSkill
- nemoclaw-autohealSkill
- nemoclaw-nvteamSkill
- outlook-email-searchSkill
- slack-channel-finderSkill
- slack-channel-summarizerSkill
- source-etl-querySkill
- module-1Skill
- module-2Skill
- module-3Skill
- module-4Skill
- module-5Skill
- module-6Skill
- module-7Skill
- setup-workshop-nemoclaw-operatorSkill
- setup-workshop-nemoclawSkill
- workshopSkill
- cross-source-gap-analysisSkill
- github-readonly-liveSkill
- gitlab-readonly-liveSkill
- nemoclaw-autohealSkill
- nemoclaw-nvteamSkill
- outlook-email-searchSkill
- public-web-searchSkill
- slack-channel-finderSkill
- source-etl-querySkill
- kubernetes-sreSkill
- openshift-llm-deploySkill
- inbound-judgingSkill
- memory-consolidationSkill
- memory-repairSkill
- memory-writingSkill
- obligation-reviewSkill
- preference-updateSkill
- nv-tech-assistantSkill
- payment-ops-playbookSkill
- payment-screeningSkill
- rail-boundary-testSkill
- release-packetSkill
- pr-reviewSkill
- pr-test-case-assistantSkill
- vss-videoSkill
- nemoclaw-hermes-swarmSkill
- blackwall-payment-gateSkill
- retail-apiSkill
- retail-apiSkill
- watchtowerSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 8 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use this skill to resolve a Slack channel and produce a summary that is bounded to retrieved history and linked to source messages.
- Distinctive: line 15
Check what was discussed in a channel before answering a question.
- Very specific: line 19
Use the provided Python helpers.
- Distinctive: line 22
Do not print the access token or include it in a command argument.
- Very specific: line 29
If the user gives a direct Slack mention like <#C0ALN454EH4>, use that ID directly.
- Distinctive: line 32
If the request comes from a tagged Slack channel and the runtime context already gives you the current channel ID, treat that as the …
- Distinctive: line 67
Read these fields before you summarize:
- Very specific: line 85
Do not summarize when ok is false.
- Distinctive: line 101
Start with a short coverage statement that gives the requested range, retrieved range, number of messages inspected, number of human …
- Distinctive: line 105
Use only retrieved message text for these sections:
- Distinctive: line 117
If there is not enough retrieved evidence for a requested conclusion, say that the retrieved messages do not establish it.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
slack-channel-summarizer
Use this skill to resolve a Slack channel and produce a summary that is bounded to retrieved history and linked to source messages.
When to use
- Summarize recent activity in a channel.
- Review conversation history for a time range.
- Check what was discussed in a channel before answering a question.
Instructions
- Use the provided Python helpers. Do not use direct
curlrequests or create another Slack client. SLACK_BOT_TOKENcontains a placeholder that OpenShell resolves on egress.- Do not print the access token or include it in a command argument.
- Summarize only messages returned by
fetch_slack_history.py.
Procedure
1. Resolve the channel ID
If the user gives a direct Slack mention like <#C0ALN454EH4>, use that ID
directly.
If the request comes from a tagged Slack channel and the runtime context already gives you the current channel ID, treat that as the resolved channel ID for phrases like "this channel".
Otherwise, use your Slack channel finder skill.
2. Read bounded channel history
For a recent-message request, run:
/usr/bin/python3 /sandbox/.hermes-data/skills/slack-channel-summarizer/scripts/fetch_slack_history.py \
--channel-id CHANNEL_ID --message-limit 10 --page-cap 10For a time range, pass inclusive --oldest and --latest boundaries as Slack
timestamps or ISO 8601 timestamps. For example:
oldest="$(date -u -d '7 days ago' +%s)"
latest="$(date -u +%s)"
/usr/bin/python3 /sandbox/.hermes-data/skills/slack-channel-summarizer/scripts/fetch_slack_history.py \
--channel-id CHANNEL_ID \
--oldest "$oldest" --latest "$latest" \
--message-limit 100 --page-cap 10 \
--replies --thread-cap 10 --reply-limit 20 --thread-page-cap 3The helper filters bot and system messages while it paginates. The limits and page caps bound the amount of channel and thread history that it returns. The hard maxima are 200 human channel messages, 25 history pages, 10 thread roots, 20 human replies per thread, and 5 pages per thread.
3. Check the evidence contract
Read these fields before you summarize:
ok: required history and source links were retrieved.empty: the completed query returned no matching human messages.coverage.requested_range: boundaries sent to Slack.coverage.retrieved_range: oldest and latest messages inspected.coverage.inspected_messages,coverage.human_messages, andcoverage.pages: the amount of history processed.coverage.completeandcoverage.truncation_reasons: whether the history query reached its boundary without hitting a cap.threads.completeandthreads.truncation_reasons: whether requested thread expansion reached its bounds.messages[].timestamp,messages[].permalink, andmessages[].citation: source metadata for each message.messages[].text_truncated: whether the helper capped that message's text.messages[].thread_replies[]: bounded replies. Each reply includesthread_root_tsso that its relationship to the root is explicit.
Do not summarize when ok is false. Report the returned stage, error, and
actionable fields instead. Common errors include:
invalid_auth: check the configured bot access token.not_in_channel: invite the bot to the channel.missing_scope: add the scope named inneeded, then reinstall the app.rate_limited: wait for the returnedretry_afterinterval.channel_not_found: check the channel ID and the bot's access.
An ok: true result with empty: true and coverage.complete: true means that
the requested range contains no matching human messages. If
coverage.complete is false, disclose the truncation reasons. Do not describe
an incomplete empty result as no channel activity.
4. Write the grounded summary
Start with a short coverage statement that gives the requested range, retrieved range, number of messages inspected, number of human messages used, and any history or thread truncation.
Use only retrieved message text for these sections:
- Main topics
- Active participants
- Decisions and action items
- Unresolved questions
Copy the citation value at the end of every factual bullet about a theme,
decision, action item, or unresolved question. Cite more than one representative
message when a conclusion combines several messages. A citation is a Slack
permalink and preserves the workspace's existing access controls.
If there is not enough retrieved evidence for a requested conclusion, say that the retrieved messages do not establish it. Do not add general patterns, background assumptions, or conclusions from channel metadata alone.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

