Use Cases

Documentation Index

Fetch the complete documentation index at: https://developer.copera.ai/llms.txt

Use this file to discover all available pages before exploring further.

Copera’s surface is intentionally generic — boards for structured data, docs for prose, drive for files, channels for messaging. The CLI skill teaches an agent the commands; what to do with them is up to you. This page collects the patterns we see most often. None of these are built-in features — each is a workflow you describe to the agent, and (for the recurring ones) capture as a workflow skill so any future session runs it identically.

---

​

Boards: track work where your team already lives

​

Triage incoming tasks

You drop in a Sentry link, a customer email, or a paste from Slack. The agent files it as a row in your tracker board with severity inferred, status set to Triage, and (for high-severity) a notification posted to your alerts channel. What the agent does

• Reads the schema of your tracker table once (cached).
• Maps your free-text input → the right column values.
• Creates the row and (conditionally) posts a notification.

Build this with a workflow skill like triage-bug or triage-task. The agent’s interview captures which board + table you use, which severity options exist, and your team’s notification channel.

​

Sales engineering pipeline

A board with a row per inbound technical request: which customer, which product area, what’s blocking the deal, who owns the response. The agent files new requests as you describe them, updates ownership/status as you go, and surfaces stale rows in your weekly review. What the agent does

• New request → rows create with customer LINK column resolved (looks up the customer row by name).
• Status change → rows update for the relevant column.
• Weekly review → rows list filtered by stale rows, summarized in a channel post or generated doc.

Tip: capture this as two separate workflow skills (sales-eng-intake and sales-eng-review) so the recurring weekly summary doesn’t drag along the schema overhead of intake.

​

Customer onboarding

When a deal closes, the agent creates the customer’s CRM row, links contacts, generates a kickoff doc from a template, drops it under the right parent, and posts a “kickoff scheduled” message to your customer-success channel. What the agent does

• rows create in CRM with LINK columns to contact rows.
• docs create under the customer-onboarding parent doc, content from a template.
• channels message send with the doc URL.

​

Status reporting

End of week: agent queries rows in your engineering tracker that were updated this week, groups by owner or by status, and posts a digest to your team channel. Or generates a doc that lives under /Engineering/Weekly/. What the agent does

• rows list (paginated if needed).
• Group + summarize in plain text.
• channels message send or docs create.

---

​

Docs: Copera as your agent’s knowledge base

Copera Docs is a workspace-wide markdown tree. That makes it a natural fit for AI agents — durable, searchable, version-controlled prose that any agent can read for context.

​

Read for context (knowledge retrieval)

You ask: “What’s our policy on customer-data deletion?” The agent searches the doc tree, reads the matching doc(s), and answers from the prose — citing the doc URL so you can verify. What the agent does

• docs search "customer data deletion" --json → finds matching docs.
• docs content <docId> for the top hit (cached) → reads the markdown.
• Answers from the doc, citing the source.

Why this works: the docs cache means follow-up questions on the same topic don’t re-fetch. And because the prose is yours, the agent stops guessing from training data.

​

Weekly meeting summaries

You keep meeting notes under /Engineering/Meetings/ (or wherever). The agent reads last week’s meeting docs, extracts decisions and action items, and posts the summary back as a new doc under /Engineering/Weekly-Summaries/ with links to the source meetings. What the agent does

• docs tree --parent <meetings-folder> --json → list recent meeting docs.
• docs content <docId> for each (concurrent reads if many).
• Synthesize summary + action items.
• docs create --title "Week of …" --parent <summaries-folder>.
• Optionally channels message send with the summary URL.

​

Create and share new docs from a conversation

Mid-call, you tell the agent: “Document this incident — what we hit, the workaround, the owner.” The agent drafts the markdown from your conversation, creates the doc under /Engineering/Incidents/, posts the link to your incident channel, and (if you’ve set it up) adds a row to your incident-tracking board. What the agent does

• docs create --title "Incident: …" --parent <incidents-folder> --content "<markdown>".
• channels message send with the link.
• (Optional) rows create in the incidents board with the doc URL as a column.

​

Doc review and edit suggestions

You point the agent at a draft doc. It reads the content, suggests edits inline (in your conversation), and on your approval applies them with docs update --operation replace. What the agent does

• docs content <docId> to read.
• Propose changes in conversation.
• On approval, docs update <docId> with the edited markdown. Async — re-reads after a few seconds to confirm.

---

​

Drive: file workflows

​

Organize uploads by convention

You drop files into a conversation. The agent uploads them to drive under a folder structure matching your team’s convention (e.g. /<year>/<quarter>/<customer>/), creating folders as needed. What the agent does

• drive tree --parent <root> to find existing structure.
• drive mkdir for any missing folders.
• drive upload <file> --parent <resolved-folderId>.

​

Distribute reports

The agent generates a report (markdown or rendered output), uploads it to a shared drive folder, and posts the link to the relevant channel(s) — replacing the typical “screenshot the dashboard, paste into Slack” loop. What the agent does

• Generate the report locally (or as a doc + export).
• drive upload report.pdf --parent <reports-folderId>.
• channels message send with the file URL from the upload response.

---

​

Cross-surface workflows

The most useful workflow skills span multiple Copera surfaces.

​

Daily standup digest

The agent queries: rows in the sprint board updated since yesterday, recent doc activity in the team’s docs folder, drive uploads in shared assets. Synthesizes into a one-message summary posted to the standup channel before the meeting.

​

Customer touchpoint summary

For a given customer (resolved to their CRM row + linked rows): pull all their open tickets, recent meeting notes, and any drive assets attached to them. Generate a one-page markdown brief the agent can share before a customer call.

​

Incident retrospective

When you close an incident: agent reads the incident’s row in the incidents board, the incident doc, and any related rows linked from the doc. Generates a retrospective doc under /Engineering/Retros/ and links it back to the incident row’s description.

---

​

What goes in a workflow skill vs ad-hoc

Not every use case needs to become a workflow skill. Rule of thumb:

• Ad-hoc — one-off discovery, occasional reads, exploring an unfamiliar board. Just talk to the agent; the CLI skill handles it.
• Workflow skill — anything you do more than twice with the same shape. Build it once, run it identically every time, share with your team. See Workflow Skills.