See Aisman in action

A closer look at the real screens Aisman uses to track issues, prompts, AI sessions, and verification—end to end.

Dashboard

Project overview, statistics, and quick-access tools at a glance.

Dashboard

A single **overview of a project's health**: total issues, remaining work, AI sessions, steps, and lines written. See how long a project has been under development and breakdown of costs.

New AI outputs detected

Aisman detects new output files and shows their status. When ready, it lets you process those files, which results in session, session step and distinct issues automatic creation.

Multiple AI output files

Multiple output files detected simultaneously, ready for batch processing.

Processed outputs

List of previously processed output files with their results.

Output file detail

Detailed view of a single output file showing its contents and processing result.

Check-in message

Project check-in message showing the current status and next steps.

The AI cost tab

Project AI cost breakdown showing usage by user, model, and host.

Issues

Browse, filter, and organize all your issues in one unified view.

The Issues main tab

Here you'll find the list of every single issue that has been created within the current project. This tab allows you to filter and search issues and of course show full details of every issue.

Issue relationships

Issue relations show how the issue relates to other issues or how other issues relate to it. Relations connect issues across your project. An issue can depend on another, extend it, or simply relate to it. Aisman tracks these automatically so you always understand the full picture.

Children and Relations

Parent-child issue hierarchies combined with cross-cutting relations.

Issue session and step membership

Every issue links back to the AI sessions and steps that touched it, giving you a complete, searchable history of how a feature or fix actually came to be.

AI Sessions

Track every AI coding session — inputs, outputs, and file changes.

AI sessions

Every AI coding session is tracked step by step, with the tool and agent used, lines added and removed, and a running list of the issues it touched-so nothing gets lost between sessions.

Session steps overview

Overview of all steps within a session, showing the progression of work.

Session step detail

Each session step shows a description of what the AI agent was asked to do, plus the outcome notes recording what actually happened — keeping a clear audit trail.

Session step outcomes

The AI agent produces both the summary description of what work the step included and also the outcome of the implementation.

Session step outcomes 2

Another session step showing different outcome patterns and metrics.

Session step issues

Complete list of all issues touched by a session step, showing their number, title, status, and origin.

Session step unfinished issues

The session step issue list filtered to only show issues that have not yet been marked as Done.

All session issues

Session view showing issues from ALL the session steps and their current states.

Transitioned and Ready issues

Issues that have been transitioned to consolidation session(s) alongside newly ready issues.

Consolidation session

Issues that were moved from other sessions, preserving their full history.

Original input text

The original free-text instructions that were part of the prompt text sent to the AI agent, preserved for reference.

Workflow

Build structured AI prompts from your issue context automatically.

Step 1: List of unresolved session issues

All still unresolves issues in the selected session .Issues from the previous step are highlighted for quick re-selection.

Step 1: Select Issues - parent-child relationships

The selected and expanded issue is a parent of the issues below, which is clearly visible from the intentation hierarchy.

Step 1: Child issue expanded

Child issue expanded below the mostly hidden parent issue in Step 1 of the workflow. Note the complexity bar (green in this easy issue's case and red for the parent issue)

Step 1: Steering on unfinished issue

After processing, if the issue still cannot be considered fully implemented, you can add notes and comments to issues to steer the AI agent in the next step.

Step 1: New vs. carried-over issues

Shows issue recency and implementation status.

Step 1: Issue rsolution note

After processing, issues show their resolution note to inform the developer what the AI did and what it thinks it accomplished doing so.

Step 1: Issue complexity and actions

Detailed resolution view including the estimated issue implementation complexity and action buttons to resolve or steer further.

Step 2: Free-text instructions

Here you write your free-text, unstructured instructions that the AI translates into distinct inter-related issues during processing.

Step 2: Free-text insructions preview

Step 2 of the workflow lets you define fresh new requirements for your software. These are then parsed into distinct issues and implemented by your favourite ai agent.

Step 3: Prompt generation

Aisman compiles selected issues, prior notes, related issue details and project instructions into a single, ready-to-paste prompt for your AI coding agent-no manual copy-pasting of scattered context required.

Step 4: Process output

The output processing step where Aisman validates and imports the agent results.

Step 4: Output file created

Confirmation that the output file has been created and issues it contains are being worked on.

Step processing finished notification

This pop-up notification is shown to you whereever within Aisman dashboard you are. It notifies you that the AI agent has probably finished processing and your action is needed.

Step 4: Output file ready

In this example, one session step is finished and ready for parsing into Aisman, and another one's output file contains an error or warning that the AI agent may need to fix.

Step 4: An output file need fixing

When an output file fails validation, Aisman tells you about it and offers a button to generate a prompt for the AI agent to fix the file.

Targeted output file fix prompt

When the AI agent's output file contains errors, Aisman displays and copies a detailed error list to your clipboard automatically so you can see exactly what went wrong and to request a fix easily.

Step 4: Output file rrors fixed

After applying the fix prompt, the output file is re-processed successfully.

Step 4: AI output processed

Once a valid AI output file is successfully imported into Aisman you are given tools to get a Git checkin message and to view the results of the work, test, confirm and/or steer the next step.

Verification

Before results are accepted, Aisman automatically checks the agent's output file for errors and warnings, confirming the file was processed successfully with no manual review needed.

Settings

Configure workflows, statuses, and integrations for your project.

Info and ToDo

The Info tab showing project information and pending to-do items.

Host instructions

The Host instructions tab shows some o the most common supported AI coding harnesses with their default instruction file paths. It is up to you to set up the ones that you use. Each must contain a reference to the Aisman instructions.

Adding Aisman instructions

One-click insertion of the Aisman instructions header and basic coding principles into your AI agent's instruction file.

Aisman instructions template

The current latest Aisman instructions template as defined on the server.

Local Aisman Instructions

The Aisman instructions template that lives in `.aisman/instructions.md` and int the sub-files in the `/instructions` folder.

Aisman instructions update review

Review mode for updating Aisman instructions with diff highlighting.

Computer name

The current computer's name (custom, not dependent on the actual machine name) is required because Aisman tracks which machine is running each session. You are asked for it when setting up the very first new project.