# Don’t just stare at the points: what’s really interesting about these 3 early projects today
Today we’re looking at **Axis Robotics, TermiX, Intuition**: one just launched a 30-day limited mission, one ties points to real Agent work, and one has just switched scoring standards between new and old tasks.
What’s worth watching now isn’t because of “what might get paid out”—it’s because all three have, within the past 72 hours, introduced verifiable new rules or new entry points.
Even the highest risk is very specific: signing effectively but failing to complete it, using funds to create meaningless interactions, and treating old tasks with already stopped bonus payouts as new tasks.
## Axis Robotics
**One-sentence highlight:** This is not an indefinite check-in—it's a 30-day contribution campaign with daily task capacity, a review process, and a final on-chain signature.
**Why now:** The official Binance Wallet exclusive campaign starts on August 26 and runs until September 25. The independent points pool is divided into two periods; about 20,000 tasks are released per day. After users submit robot demo trajectory data, they must wait for review. Only after passing review and completing the signature on Base will an effective record be created. What’s truly scarce here isn’t the click-through entry, but the task slots, qualifying trajectories, and the signature time before the deadline.
**Metrics worth tracking:**
- Your daily task claim speed and the remaining quantity after each refresh;
- Your submission pass rate, not just the number of submissions;
- The number of items pending signature in your Portfolio, and the backlog when the deadline is approaching.
**Pitfalls I’ll avoid:** Wallets that import a mnemonic phrase do not meet the eligibility for this exclusive campaign; simply completing the trajectory without a final on-chain signature may also not be counted. Review may take up to 48 hours, and the last few days concentrate submissions—time risk is very high.
## TermiX
**One-sentence highlight:** The latest points guidelines put the “real Agent business activity” in the spotlight, while social tasks are only the shallowest layer.
**Why now:** On August 27, the official team published six types of points paths in a centralized way, including original content, registering an Agent, publishing and fulfilling tasks, delivering work, and settling with stablecoins. The most important thing to watch is Agent Jobs: the project has started distinguishing between people who merely “visited the page” and people who truly generate tasks, deliver, and settle. For content creators, the same high-quality original content can be submitted under different activity rules, but that doesn’t mean repeated spammy posting is effective.
**Metrics worth tracking:**
- Agent Job completion rate, settlement rate, and dispute rate;
- Whether the points page discloses different contribution weights;
- Whether low-quality content, batch Agents, and looping tasks are filtered out.
**Pitfalls I’ll avoid:** I won’t create fake orders between two wallets just to chase points, and I won’t treat stablecoin settlement as a zero-cost action. Task delivery involves counterparty and acceptance risks, and the currently published points guidelines cannot be used to infer future redemption relationships.
## Intuition
**One-sentence highlight:** New and old tasks have already been routed separately—adding more Season 1 old tasks will no longer increase IQ Points.
**Why now:** The official team clearly stated on August 27 that although Legacy Quests can still help establish on-chain records, they have stopped issuing IQ Points. Going forward, you should only look at new tasks that are still marked as valid in the Portal. Recently, the community has shown Preference-type tasks, but the required assets, on-chain steps, and points on the page should be based on what was applicable at the time of operation—you can’t treat screenshots as permanent fixed rules.
**Metrics worth tracking:**
- Frequency of new task launches and their validity periods;
- The actual changes in IQ Points before and after the operation;
- Whether staking or on-chain expressions can be withdrawn, and what the exit cost is.
**Pitfalls I’ll avoid:** I won’t assume that an old task still has points just because it can be opened; and I won’t buy unfamiliar assets only to chase the page numbers. The knowledge graph emphasizes information quality—choosing casually or submitting in batches may leave low-quality on-chain trails.
## Collectible Judgment Framework
When facing any points or early tasks, first pass through four gates: **whether the nodes are officially confirmed, whether actions form an effective record, whether costs can be exited at any time, and whether failure conditions are clearly stated.** Then split the metrics into three columns: investment (time, Gas, asset exposure), effectiveness (review, signature, funds received), and rule risks (eligibility, anti-sybil, deadlines). If any one of the four gates is unclear, observe first; only when all three columns can be recorded is it worth continuing to invest time.