# Review agent trajectories
Generated: 2026-11-13T08:01:29.487Z
Source build: local
Canonical docs: https://teammately.ai/docs
---
id: workflow.expert-contributions.review-trajectories
title: Review agent trajectories
summary: Assess the actions an agent took when its final answer cannot establish whether the work was acceptable.
kind: task
product_area: expert_contributions
status: stable
updated: 2026-11-11
canonical: /docs/expert-contributions/review-trajectories
---
# Review agent trajectories
Assess the actions an agent took when its final answer cannot establish whether the work was acceptable.
## Prerequisites
The review activity includes an actual recorded trajectory and the applicable task context.
## Steps
1. Read the task and identify the behavior the expert is being asked to judge.
2. Inspect recorded tool use, observations, actions, and supplied reasoning summaries. Distinguish absent records from recorded non-action.
3. Explain which observed step met or violated the requirement and what condition made it consequential.
4. Review any resulting criterion against what the recording can actually establish.
## Object and state changes
Your review stays attached to the recorded behavior and Contribution. It does not reconstruct unrecorded private reasoning or certify actions outside the captured scope.
## Success criteria
The judgment identifies the relevant recorded action and a requirement the team can apply to comparable behavior.
## Common failure modes
Inferring a correct process from a correct final answer; assuming a missing event proves an action never occurred.
## Related reference pages
{% related-card-grid title="Continue this workflow" %}
- [Contribute through forms, chat, and voice](/docs/expert-contributions/complete-contribution)
{% /related-card-grid %}
## Related troubleshooting pages
{% related-card-grid title="Resolve a problem" %}
- [Expert Contribution problems](/docs/troubleshooting/expert-contributions)
- [Low expert agreement](/docs/troubleshooting/low-expert-agreement)
{% /related-card-grid %}
---
id: expert-contributions.complete
title: Contribute through forms, chat, and voice
summary: Explain your judgment, review Lemon’s interpretation, and complete the assigned contribution through discussion and focused review.
kind: task
product_area: expert_contributions
status: stable
updated: 2026-11-11
canonical: /docs/expert-contributions/complete-contribution
---
# Contribute through forms, chat, and voice
Your expertise helps the team understand what good behavior requires and why. Read the assignment, examine the prepared material, and explain the condition, exception, or tradeoff behind your judgment.
Lemon can follow up on your answers and bring proposed interpretations back for review.
## Prerequisites
- Your authenticated Contribution link and assigned work.
- The Case, reference material, or question you are being asked to examine.
- Microphone permission for a voice interview.
## Steps
1. Open the Contribution and read its objective and assignment context.
2. Examine the Case, outputs, or question presented. Follow the relevant documents and attachments when they affect your decision.
3. Use the review, comparison, form, conversation, or interview controls to explain your judgment.
4. Point out a missing condition or uncertainty when it changes the answer.
5. Review proposed policies, applicability, and rubrics. Use their decision controls to accept, reject, or correct the interpretation.
6. Continue through the assigned work and its final review. Read the completion message for the Contribution.
## Work through a conversation or a focused review
In a conversational Contribution, Lemon brings questions and prepared activities into the discussion. Open the presented review or interview when it helps examine the distinction. You can return to completed activities and earlier messages for context.
Focused reviews let you inspect a Case, curate an answer, compare alternatives, or assess a recorded trajectory. Explain why the choice matters so the team can apply your reasoning beyond that example.
## Review the proposed standard
Read the exact wording and the Cases to which it applies. Correct an omitted exception or a scope that changes your meaning. Discussion develops the interpretation; the artifact’s approval or rejection control records your decision on that proposal.
Final policy review brings the resulting interpretation back to you before the Contribution moves through its remaining rubric and completion work.
## Continue smoothly through the assignment
Next and Skip move to prepared work while saved answers are delivered in the background. If delivery needs attention, retain your answer and follow the displayed recovery guidance.
Finalization preserves submitted work while the Contribution settles its remaining decisions. The completion message identifies when the assignment is finished.
## Object and state changes
Your responses, rationale, and artifact decisions stay attached to the Contribution and the content reviewed. Accepted findings can inform policies, rubrics, Cases, and coverage through their owning workflows.
## Success criteria
- Your reasoning and important qualifications are clear.
- Proposed standards preserve your intent and applicability.
- Explicit artifact decisions reflect the wording you reviewed.
- The Contribution reaches its completion state.
## Common failure modes
- Choosing an answer without explaining the condition behind it.
- Accepting wording that omits a consequential exception.
- Using conversational agreement in place of the proposal’s decision control.
- Re-entering a different answer while an earlier submission is being saved.
{% example-demo title="Qualify an exception" %}
Lemon proposes that expired agreements should always be ignored. You explain that an older agreement can remain controlling when the current contract incorporates it. Review the revised policy and its applicable Cases, then record your decision on the corrected interpretation.
{% /example-demo %}
## Related reference pages
{% related-card-grid title="Understand your contribution" %}
- [Expert UI](/docs/integrations/reviewer-workspace)
- [Contributed Artifacts](/docs/expert-contributions/contributed-artifacts)
- [Contribution Lifecycle and Status](/docs/expert-contributions/lifecycle-and-status)
{% /related-card-grid %}
## Related troubleshooting pages
{% related-card-grid title="Resume your work" %}
- [Expert Contribution Problems](/docs/troubleshooting/expert-contributions)
- [Permissions](/docs/troubleshooting/permissions)
{% /related-card-grid %}
---
id: troubleshooting.expert-contributions
title: Expert Contribution problems
summary: Diagnose contribution access, task routing, checkpoint, synchronization, completion, and artifact-reconciliation problems.
kind: error
product_area: troubleshooting
status: stable
updated: 2026-11-11
canonical: /docs/troubleshooting/expert-contributions
---
# Expert Contribution problems
## Symptom
An expert cannot open the Contribution, reaches the wrong task, sees stale work, cannot submit a response, remains at a checkpoint or waiting state, or completes the experience without the administrator seeing reconciled artifacts.
## Likely causes
- The expert lacks access or opened a link for a different Contribution.
- The task handoff changed and the browser still points to an earlier route.
- A form, chat, interview, or case-review command has not synchronized.
- The task lease, realtime connection, or retry state needs recovery.
- A checkpoint still requires confirmation or correction.
- The expert experience is complete, but contributed artifacts have not yet been reconciled in the administrator workspace.
- The Contribution was cancelled or changed while the expert was working.
## Fix
1. Confirm the Contribution identity and expert assignment in **Expert Contributions → Contributions**.
2. Inspect its current status and task plan rather than relying on the expert's previous URL.
3. Distinguish an outdated bookmark from pending answer delivery. Use the current Contribution entry for an outdated bookmark only after staged and unsent input is protected.
4. Ordinary submission, activation and synchronization recover in the background. Preserve exact pending input; do not repeatedly submit, reload or navigate to make delivery progress. A genuine takeover or revoked assignment remains a separate authority decision.
5. Inspect checkpoints for an unconfirmed or corrected statement.
6. Use **Logs & Status** to distinguish expert completion from reconciliation or artifact-materialization work.
7. Check **Contributed Artifacts** and the owning Correctness Governance, Assets, or Coverage Design surface before concluding that accepted material is missing.
## Prevention
- Keep Contribution objectives and task sequences narrow.
- Verify expert access before sending the request.
- Preserve checkpoints for consequential meaning.
- Use the current Contribution entry route instead of bookmarking an internal task route.
- Teach administrators that task completion, artifact reconciliation, and artifact governance are separate transitions.
- Avoid submitting the same response repeatedly when the product shows a pending or retryable synchronization state.
## Related task pages
{% related-card-grid title="Related task pages" %}
- [Request an Expert Contribution](/docs/expert-contributions/request-contribution)
- [Complete an Expert Contribution](/docs/expert-contributions/complete-contribution)
- [Complete your first alignment loop](/docs/quickstart)
{% /related-card-grid %}
## Related reference pages
{% related-card-grid title="Related reference pages" %}
- [Expert Contributions](/docs/expert-contributions)
- [Contributed Artifacts](/docs/expert-contributions/contributed-artifacts)
- [Permissions](/docs/reference/permissions)
{% /related-card-grid %}
---
id: troubleshooting.low-expert-agreement
title: Low expert agreement
summary: Use disagreement to clarify context, applicability, source authority, and unresolved product policy.
kind: error
product_area: troubleshooting
status: stable
updated: 2026-11-06
canonical: /docs/troubleshooting/low-expert-agreement
---
# Low expert agreement
## Symptoms
- Experts choose different answers for the same Case or Checkpoint.
- Rationale shows different assumptions even when the selected answer matches.
- Agreement is high on routine Cases and collapses on one boundary segment.
- A Contribution completes, but its learning cannot be materialized into one defensible standard.
## Likely causes
- Experts saw different Case versions, sources, or task context.
- The Case or question leaves the required judgment ambiguous.
- Reviewers use different applicability or source-authority assumptions.
- The domain genuinely contains unresolved or context-dependent policy.
## Preserve disagreement before interpreting it
Do not average away the responses or select the majority answer before reading rationale. Confirm that each response belongs to the intended expert, task, Case version, question, and source context. A routing mistake and a legitimate domain disagreement require different fixes.
## Diagnostic checks
1. Open the review or session detail and compare the exact prompt, Case context, available sources, response, rationale, and timestamps.
2. Check whether all experts saw the same version and whether any required context was missing.
3. Separate four causes: ambiguous Case, unclear question, different applicability assumptions, and conflicting domain judgment.
4. Segment the disagreement by source authority, user type, risk, product state, or other relevant coverage facet.
5. Ask whether the current Policy already resolves the question. If it does, determine whether the Policy or its applicability was invisible or unclear to reviewers.
## Fix
- **Missing context:** correct or version the Case and repeat only the affected review.
- **Unclear question:** rewrite the task so the requested judgment and required source context are explicit.
- **Different applicability:** refine the Policy boundary or split Cases into distinct contexts.
- **Conflicting expert judgment:** route the findings through conflict resolution and an accountable owner.
- **Wrong expertise:** revise assignment criteria without deleting the attributable response.
Do not treat consensus as approval. Even unanimous expert input becomes a governed Policy, Rubric, Case-scoped reference output, or Benchmark Dataset decision only through the owning workflow.
## Prevention
Package the same visible context for every assigned expert, state the requested judgment precisely, and include applicability or source-authority context when it controls the answer. Preserve rationale so later disagreement can be diagnosed without repeating every Contribution.
## Related task pages
{% related-card-grid title="Related workflows" %}
- [Resolve conflicting correctness results](/docs/governance/conflict-resolution)
- [Complete an Expert Contribution](/docs/expert-contributions/complete-contribution)
- [Boundary Cases](/docs/coverage-engineering/boundary-cases)
- [Approval History and Reviewer Activity](/docs/governance/approval-history-and-reviewer-activity)
{% /related-card-grid %}
## Related reference pages
{% related-card-grid title="Related reference" %}
- [Expert Contributions](/docs/expert-contributions)
- [Human Approval Boundaries](/docs/governance/human-approval-boundaries)
- [Reviewer and Project Access](/docs/governance/reviewer-and-project-access)
{% /related-card-grid %}