Quests

Quests are the primary task units in QuestLore. Each quest represents a skill-building activity, project, or challenge that employees can undertake. When a member completes a quest and it is approved by a reviewer, they earn Experience Points and Coins.

Managers and owners create quests in the Quest Management area (Organization → Quests). Employees discover and accept available quests from the Quests Library.

Every quest has a difficulty level that determines how much Experience Points and Coins it awards upon successful completion. There are four difficulty tiers:

• Easy — Short, straightforward tasks suitable for onboarding or routine work. Experience Points and Coin rewards are lower.
• Medium — Moderately complex tasks that require some effort or skill. Rewards are proportionally higher.
• Hard — Challenging tasks demanding significant expertise or time. Yield substantially more Experience Points and Coins.
• Expert — The highest tier. Reserved for the most demanding, high-impact work. Expert quests carry the largest rewards.

The exact Coin reward for each difficulty tier is set org-wide by the owner under Organization → Settings → Rewards Economy. Every quest of the same difficulty pays the same number of Coins — there is no per-quest coin override.

A quest moves through three statuses during its lifetime:

• Draft — The quest has been created but is not yet visible to employees. Managers use this status while configuring quest details.
• Active — The quest is live and available in the Quests Library for employees to accept and work on.
• Archived — The quest is no longer available for new assignments. Historical completions are preserved.

Quests can optionally be configured as repeatable. A repeatable quest can be accepted and completed by the same member more than once, subject to a cooldown period set by the manager.

A cooldown value (in hours) is required for all repeatable quests. Once a reviewer approves or rejects a submission, the cooldown timer starts. The member cannot start a new attempt until the cooldown expires. Time spent waiting in the review queue does not count against the cooldown — the clock only begins after a final decision is made.

Non-repeatable quests can only be completed once per member. Once approved, the quest no longer appears as available for that member.

Quests can be tagged with one or more categories (for example, "Leadership", "Technical Skills", "Customer Success"). Categories help employees filter the Quests Library and help managers organise their quest catalogue.

Categories are created and managed by owners and managers under Organization → Categories and can be shared across quests and products.

A quest's visibility controls who can see and accept it:

• Public — Visible to all active employees in the organisation's Quests Library.
• Private — Only visible and assignable to specific members or departments selected by the manager at the time of assignment.

Every quest goes through a single review pipeline:

1. AI review first — When a member submits a quest, the AI Reviewer evaluates the work against the quest's Acceptance Criteria. It reads the written submission, attachments (images, PDFs, Word and Excel documents, and text files), and — when the submitter has connected their GitHub account — the description and code diff of any linked pull request. If the AI is confident in its decision, it can approve, reject, or request changes immediately.
2. Escalation to a human — The system assigns a human reviewer when the AI is not confident in its decision, encounters content it cannot evaluate (such as an unsupported file type or a resource it cannot access), or has already requested changes twice. From that point on the human is the sole reviewer for the rest of the quest's life.
3. Manager-assigned quests skip AI — When a manager or owner uses "Assign quest to member", the assigning manager becomes the sole reviewer immediately. AI is bypassed and the submitter sees the manager's identity from the start.

Acceptance Criteria are required for every quest. They are the binding requirements the AI scores against — keep them concrete and verifiable.

See the Peer Review article for more on how human reviewers are assigned and how reviewer identity is anonymised to the submitter until a final decision is posted.