01 · Pick the work
Stories · Search
Approved
- StoryP13 ptsApproved
Save search filters
- BugP21 ptsApproved
Filters reset on back
Workstream
Saved search filters
Use case
Approved work usually sits in a tracker until someone copies it into a coding agent by hand, and leaves half the context behind. In Spectr, a developer groups the stories and tasks into a workstream and starts a Conductor run from it. Spectr writes the brief from the stories themselves, and the run's progress shows on the board.
How it works
Two approved stories become one workstream on the Development board. The developer starts a Conductor run from it: their own Conductor account, the repository matched to the project's GitHub repo, and the stories to send. Spectr shows the exact message and sends it only on confirm. The workstream moves to In Progress and the run's state shows on its card. The run uses the developer's Conductor compute, not the workspace's Spectr AI allowance.
A coding agent works from whatever it is given. Paste in a ticket title and it fills the gaps with guesses: a constraint from the call gets dropped, a design gets ignored, an edge case nobody wrote down gets built wrong. The spec existed. It never reached the agent.
Two developers. Three approved stories for saved search. Conductor connected.
Old way
Each developer copies a ticket into their agent, adds what they remember from the call, and keeps the design open in another tab.
9:10am
Now: the three stories go into one workstream on the Development board. It starts in Pending.
9:12am
Start with Conductor. The repository is already picked, because it matches the project's GitHub repo. Spectr shows the brief: each story with its acceptance criteria, links and Figma screens.
9:13am
The brief is sent. The workstream moves to In Progress, and the run on its card goes from Starting to Ready.
2:40pm
A fourth story is approved. It goes to the same run with Send stories, as one more message.
Friday
The workstream is in review. Anyone who opens the board sees who started the run and where it stands.
Instead of retyping three tickets into an agent from memory, the developer read one brief and sent it. The agent started with the acceptance criteria and designs the team had agreed.
Group related stories from one project into a workstream and move it through Pending, In Progress, In Review and Done. A story belongs to one active workstream at a time, so two people do not build the same thing.
Spectr builds one message from the stories you pick: title, type, priority, estimate, description, acceptance criteria, links and Figma screens, plus a note if you add one. You see the exact text, and nothing reaches Conductor until you confirm.
Each developer connects their own Conductor account. Runs use that account's compute, so they spend none of the workspace's Spectr AI allowance, and Development is free on every plan.
Cursor, Codex or Claude Code connected over MCP can read the story they are implementing, as the developer, while they work.
Anything else? Email hello@crowdlinker.com and a person will answer.
Conductor today. Claude, Codex and Cursor connect over MCP to read your Spectr data, and are planned as coding-agent runs later.
No. Development and connected accounts are free on every plan, and runs use your own Conductor compute, not your Spectr AI allowance.
You can untick stories and add a note, which goes after the stories as you wrote it. To change a story itself, edit it in Spectr and the brief is rebuilt from it.
Yes. Each starts runs on their own Conductor account, and everyone sees every run, who started it and its state. Only the person who started a run can send more stories to it.