tessl-eval-demo

Use when a well-defined engineering task must be carried from approved requirements to a verified, reviewed, PR-ready local change.

Use when a well-defined engineering task must be carried from approved requirements to a verified, reviewed, PR-ready local change.

Carries a code change through to a tested, verified and reviewed local change, loading the stage skill for each step: planning, test-first implementation, debugging, verification and review. Use when a request asks for code to change, whether it arrives as an approved ticket or a one-line message: a bug fix, a failing test or crash, a small feature or flag, a behavior change, or review comments to fix. Not for questions about the code, for merging or deleting a finished branch, or for deciding which review comments to accept without changing code. Use it whenever the work will change code, even when the user does not mention tests, review or a pull request.

Executes a written multi-step implementation plan end to end - loads the plan, reviews it critically for gaps, works through each task with its verification steps, and reports when everything is done. This is the execution stage inside the delivery-flow workflow, entered by requests like "implement this plan", "run the plan", "execute the plan", or "work through these tasks" once delivery-flow has produced a written plan. Do not activate directly for a standalone request that has not gone through delivery-flow first.

Executes a written multi-step implementation plan end to end - loads the plan, reviews it critically for gaps, works through each task with its verification steps, and reports when everything is done. This is the execution stage inside the delivery-flow workflow, entered by requests like "implement this plan", "run the plan", "execute the plan", or "work through these tasks" once delivery-flow has produced a written plan. Do not activate directly for a standalone request that has not gone through delivery-flow first.

Executes a written multi-step implementation plan end to end - loads the plan, reviews it critically for gaps, works through each task with its verification steps, and reports when everything is done. This is the execution stage inside the delivery-flow workflow, entered by requests like "implement this plan", "run the plan", "execute the plan", or "work through these tasks" once delivery-flow has produced a written plan. Do not activate directly for a standalone request that has not gone through delivery-flow first.

Verify tests pass, then present the integration options for a finished branch — merge locally, push and open a pull request, or keep it as-is — execute the choice, and clean up the worktree safely. This is the internal integration stage of the delivery-flow workflow. Use it only after implementation is verified and only when the user has authorized an integration decision, for example by saying "finish this branch", "merge my work", or "open a PR" — never activate it directly for an unrelated standalone request.

Verify tests pass, then present the integration options for a finished branch — merge locally, push and open a pull request, or keep it as-is — execute the choice, and clean up the worktree safely. This is the internal integration stage of the delivery-flow workflow. Use it only after implementation is verified and only when the user has authorized an integration decision, for example by saying "finish this branch", "merge my work", or "open a PR" — never activate it directly for an unrelated standalone request.

Use when the user wants to finish a development branch or asks what to do with finished work (merge it, push it and open a pull request, keep it for later, throw it away, or remove its worktree), and before any merge, push, branch deletion or worktree removal at the end of a piece of work. Runs the tests first, presents the options as the user's decision, and never destroys unmerged commits or uncommitted files without explicit confirmation.

Evaluate code review feedback on its technical merits, verify each suggestion against the codebase, then implement the fixes that hold up and push back with reasoning on the ones that don't. This is the internal review-fix stage of the delivery-flow workflow, run when delivery-flow receives feedback from a reviewer. It is not a standalone entry point, so do not activate it directly for a one-off "address this review comment" or "respond to PR feedback" request; use delivery-flow for those instead.

Evaluate code review feedback on its technical merits, verify each suggestion against the codebase, then implement the fixes that hold up and push back with reasoning on the ones that don't. This is the internal review-fix stage of the delivery-flow workflow, run when delivery-flow receives feedback from a reviewer. It is not a standalone entry point, so do not activate it directly for a one-off "address this review comment" or "respond to PR feedback" request; use delivery-flow for those instead.

Evaluate code review feedback on its technical merits, verify each suggestion against the codebase, then implement the fixes that hold up and push back with reasoning on the ones that don't. This is the internal review-fix stage of the delivery-flow workflow, run when delivery-flow receives feedback from a reviewer. It is not a standalone entry point, so do not activate it directly for a one-off "address this review comment" or "respond to PR feedback" request; use delivery-flow for those instead.

Dispatch a code-reviewer subagent with precisely scoped context (git SHAs, task description, requirements) to review a diff and return Critical, Important, and Minor findings. Use after delivery-flow finishes implementing and verifying a task, before merging to main. This is the internal code-review stage of the delivery-flow workflow. It is not a standalone entry point, so do not activate it directly for a one-off "review this PR" or "review my code" request; use delivery-flow for those instead.

Dispatch a code-reviewer subagent with precisely scoped context (git SHAs, task description, requirements) to review a diff and return Critical, Important, and Minor findings. Use after delivery-flow finishes implementing and verifying a task, before merging to main. This is the internal code-review stage of the delivery-flow workflow. It is not a standalone entry point, so do not activate it directly for a one-off "review this PR" or "review my code" request; use delivery-flow for those instead.

Dispatch a code-reviewer subagent with precisely scoped context (git SHAs, task description, requirements) to review a diff and return Critical, Important, and Minor findings. Use after delivery-flow finishes implementing and verifying a task, before merging to main. This is the internal code-review stage of the delivery-flow workflow. It is not a standalone entry point, so do not activate it directly for a one-off "review this PR" or "review my code" request; use delivery-flow for those instead.

Investigates a test failure, bug, or unexpected behavior to find its root cause before proposing any fix - reproduce the issue, trace the data flow, form and test one hypothesis at a time, then implement and verify the fix. This is the debugging stage inside the delivery-flow workflow, entered by requests like "debug this failure", "why is this test failing", "find the root cause", or "this bug keeps coming back" once delivery-flow has handed off a bug. Do not activate directly for a standalone bug report that has not gone through delivery-flow first.

Investigates a test failure, bug, or unexpected behavior to find its root cause before proposing any fix - reproduce the issue, trace the data flow, form and test one hypothesis at a time, then implement and verify the fix. This is the debugging stage inside the delivery-flow workflow, entered by requests like "debug this failure", "why is this test failing", "find the root cause", or "this bug keeps coming back" once delivery-flow has handed off a bug. Do not activate directly for a standalone bug report that has not gone through delivery-flow first.

Investigates a test failure, bug, or unexpected behavior to find its root cause before proposing any fix - reproduce the issue, trace the data flow, form and test one hypothesis at a time, then implement and verify the fix. This is the debugging stage inside the delivery-flow workflow, entered by requests like "debug this failure", "why is this test failing", "find the root cause", or "this bug keeps coming back" once delivery-flow has handed off a bug. Do not activate directly for a standalone bug report that has not gone through delivery-flow first.

Implements a feature or bug fix test-first - write one failing test, watch it fail, write the minimal code to pass, then refactor with tests green. This is the implementation stage inside the delivery-flow workflow, entered by requests like "implement this feature", "fix this bug", "write the code for this", or "add the tests and implement it" once delivery-flow is driving the change. Do not activate directly for a standalone request that has not gone through delivery-flow first.

Implements a feature or bug fix test-first - write one failing test, watch it fail, write the minimal code to pass, then refactor with tests green. This is the implementation stage inside the delivery-flow workflow, entered by requests like "implement this feature", "fix this bug", "write the code for this", or "add the tests and implement it" once delivery-flow is driving the change. Do not activate directly for a standalone request that has not gone through delivery-flow first.