# AX Check: skillprompting.com Checked 2026-11-02. Homepage is agent-readable, but pricing and docs trail off fast 6 of 23 checklist items pass, 15 are unassessed. Pricing pages and install/docs links were never reached, leaving most of onboarding and activation unverified. ## End-to-end onboarding not demonstrated ## Coding sessions All three sessions completed. DeepSeek V4.1 Flash and Kimi K3 both found a live entry fee of 0.009 SOL against a 0.06 SOL pot and flagged that USD value depends on SOL price, plus free entry paths. Qwen 3.8 Max could not find pricing at all and said it needed documentation to explain costs. ### DeepSeek V4.1 Flash [View public run](https://agents.withgauge.com/p/runs/5aed5c9e-0d56-4bd5-a24f-8faf47abadd8) · [Read transcript](https://www.ax-check.com/skillprompting.com/sessions/deepseek.json) Quoted 0.009 SOL per-entry fee and 0.06 SOL pot from live /state for the current round, noting free paths, per-entry basis, SOL volatility, and W-9 threshold. (seq 36, seq 43, seq 52, seq 54) #### End-to-end onboarding - **End-to-end onboarding not demonstrated**: No credentials were obtained and no authenticated write operation was performed. The agent only hit unauthenticated read endpoints (/eligibility, /state, /rounds, /health) and wrote a local read-only example script. It explicitly chose not to pursue the paid entry path (needs a funded Solana wallet it did not have) or the free puzzle path (needs a multi-minute CPU proof-of-work it was told to avoid), so no submission, wallet, or JWT was ever acquired or used. Event 48: ```text console.log("\n== how an entry is submitted (not executed) =="); ``` Event 54: ```text The paid path needs a funded Solana wallet, which I don't have. ``` Event 54: ```text The free puzzle path requires a multi-minute CPU proof-of-work (60s–5min), i.e. a long-running command you told me not to wait on. ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No funded Solana wallet available to pay entry fee**: The paid entry path requires a funded Solana keypair to sign a transaction before POST /submit. This is a test-environment limitation (no wallet provisioned), not a product defect — the product correctly documents this requirement. Event 54: ```text The paid path needs a funded Solana wallet, which I don't have. ``` - **Free entry path requires a long-running CPU proof-of-work**: The puzzle-based free entry path requires solving an argon2id proof-of-work that the docs say takes 60-120 seconds, up to 5 minutes. The agent was instructed not to run long-running commands, so it stopped rather than attempt the puzzle solve. This is a session constraint, not a product failure, though it does mean the only truly free, scriptable entry path is CPU-intensive by design. Event 54: ```text The free puzzle path requires a multi-minute CPU proof-of-work (60s–5min), i.e. a long-running command you told me not to wait on. ``` Event 25: ```text Typical solve time is 60–120 seconds on a 2024-era laptop core. Budget up to 5 minutes under memory contention. ``` - **MCP/Eliza quickstart requires a local running process**: The documented easiest integration paths (MCP server, Eliza plugin) both run as a local long-lived process rather than a pure hosted SDK/API call, which conflicted with the session's instruction to avoid starting local service stacks. Event 54: ```text The MCP/Eliza quickstarts require running a local process, which you also excluded. ``` #### Suggested Changes - **Add a stateless/fast free-entry option alongside the puzzle path**: The only free, fully-programmatic path documented in llms.txt is the argon2id puzzle, which takes 60 seconds to 5 minutes of CPU time per docs. Consider documenting whether a lower-difficulty or instant free-entry tier exists for quick API trials, verifiable by timing a /puzzle/nonce -> solve -> /submit round trip under 10 seconds. Event 25: ```text Typical solve time is 60–120 seconds on a 2024-era laptop core. Budget up to 5 minutes under memory contention. ``` ### Kimi K3 [View public run](https://agents.withgauge.com/p/runs/aab85402-6bbc-4dc9-96d4-1437e832d7c9) · [Read transcript](https://www.ax-check.com/skillprompting.com/sessions/kimi.json) Quoted a per-entry fee of 0.009 SOL against a 0.06 SOL pot, read from the live /state response. Noted the fee is set per round, the USD figure depends on the SOL price, and there are free puzzle and postcard entry options (seq 45, seq 25, seq 47) #### End-to-end onboarding - **End-to-end onboarding not demonstrated**: The agent only used unauthenticated, read-only endpoints (/eligibility, /state, /rounds, /round/:id/status, /round/:id/recent-entries, /round/:id/results). No credentials (API key, wallet keypair, puzzle solve, or signed transaction) were obtained or used, and no write/authenticated operation (submitting a prompt, paying an entry fee, solving the puzzle) was performed. The agent explicitly chose to stop short of any submission flow. Event 47: ```text I deliberately stopped at submission: the paid path needs a signed Solana transaction, and the free puzzle path runs argon2id proof-of-work for 1–5 minutes of CPU — the kind of long-running step you told me to avoid. ``` Event 39: ```text [1] Eligible: True (Your location is eligible) [2] Active rounds: 4 No open rounds right now. ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No open round available to demo a live submission**: The only currently active round was in 'pending' state (not yet open), so the example script could not demonstrate a live entry; the agent fell back to showing archived/completed round data instead. This is a product/timing limitation (rounds run on a schedule) rather than a credential or agent error. Event 45: ```text DAILY-20261102-f442 pending 0.06 SOL pot, 0.009 SOL fee, opens 2026-11-02T16:00:00.000Z closes 2026-11-03T01:00:00.000Z ``` Event 39: ```text [2] Active rounds: 4 No open rounds right now. ``` - **Submission paths require either a Solana wallet or a multi-minute CPU puzzle**: Entering a contest (the only authenticated/write action in the product) requires either signing a Solana transaction with a private key or solving an argon2id proof-of-work puzzle that takes 1-5 minutes of CPU time. Per the session's constraints against long-running steps, the agent chose not to pursue either path, so no write operation was ever attempted. Event 25: ```text Signing transactions requires a Solana keypair. If you're an autonomous agent, you'll need access to a private key and `@solana/web3.js`. ``` Event 47: ```text the free puzzle path runs argon2id proof-of-work for 1–5 minutes of CPU — the kind of long-running step you told me to avoid ``` #### Suggested Changes - **Normalize round ID casing across endpoints**: The agent observed a mismatch between lowercase round IDs documented in examples (e.g. daily-20260318-k7xm) and uppercase IDs actually returned by the live API (e.g. DAILY-20261102-f442). Standardize casing in both the live /state and /rounds responses and the llms.txt documentation examples, then confirm consistency by comparing a fresh GET /state response id field against the llms.txt example format. Event 47: ```text a quirky lowercase-vs-uppercase round ID situation I noticed between endpoints ``` Event 45: ```text DAILY-20261102-f442 pending 0.06 SOL pot, 0.009 SOL fee, opens 2026-11-02T16:00:00.000Z closes 2026-11-03T01:00:00.000Z ``` ### Qwen 3.8 Max [View public run](https://agents.withgauge.com/p/runs/4213c028-3255-40d9-8c59-53b4c49c45bb) · [Read transcript](https://www.ax-check.com/skillprompting.com/sessions/qwen.json) The agent gave no cost figure or assumptions. It said it couldn't explain pricing without docs. (seq 12) #### End-to-end onboarding - **End-to-end onboarding not demonstrated**: The agent never attempted to find or use product credentials. It inspected an empty local repository, found no references to the product, and immediately concluded no developer workflow existed, stopping without searching docs, trying an SDK, or hitting any API. No credentials were obtained and no authenticated operation was attempted. Event 12: ```text **"Skillprompting" doesn't appear to exist as a product I can act on.** ``` Event 12: ```text I have no knowledge of a hosted product, SDK, API, docs site, or quickstart by that name ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **Agent had no information about the named product**: The session provided only a generic prompt naming 'Skillprompting' with no docs link, SDK name, or API reference, and the local repo was empty. The agent had no web access to search for the product and treated its own lack of prior knowledge as proof the product has no usable developer workflow. This is a test-environment/setup limitation (missing starter materials) combined with the agent's inability to browse, not a demonstrated product defect. Event 5: ```text total 16 drwxr-xr-x 3 root root 4096 Sep 18 08:00 . drwxr-xr-x 7 root root 4096 Nov 2 07:04 .. drwxr-xr-x 8 root root 4096 Sep 24 20:05 .git -rw-r--r-- 1 root root 408 Sep 18 08:00 .gitignore ``` Event 6: ```text (no output) ``` Event 12: ```text I'd need you to paste the relevant docs or SDK reference here ``` #### Suggested Changes None identified in this transcript. ### Task given to each agent Help me build a simple example using Skillprompting. Tell me how pricing works, and briefly tell me whether this product will be easy for you to manage. Let me know if you get blocked. If this product has no developer workflow you can act on, say so plainly and stop. Stay light: use the hosted product through its SDK or API. Do not start local service stacks or wait for long-running commands; if the quickstart requires either, say so plainly and stop. No product credentials were supplied and no purchases were authorized. ## Score: B · 73/100 (provisional) Grades come from completed site checks. Coding sessions and skipped checks do not affect the score. ### Clarity - **Failed** — llms.txt provides an actionable documentation index ```text llms.txt request returned HTML, not a plain-text index; no actionable doc index observed. ``` - **Pass** — Homepage answers Markdown requests ```text Homepage served text/markdown via Accept: text/markdown; index.md returned 200 markdown. ``` - **Skipped** — llms.txt provides navigation guidance ```text llms.txt body was HTML, so navigation guidance could not be assessed. ``` - **Skipped** — llms.txt mentions offered API, MCP, and skills ```text llms.txt body was HTML; API/MCP/skills mentions could not be assessed. ``` - **Pass** — A compact guide representation exists ```text Homepage served as text/markdown at /index.md; site states every page available as Markdown. ``` - **Pass** — A focused guide is directly retrievable ```text Compact Markdown homepage guide retrieved directly with concrete entry, rules and agent instructions. ``` - **Pass** — Equivalent instructions fit a token budget ```text Markdown guide measured 753 tokens, well under the 8000-token budget. ``` - **Pass** — Product-docs links survive format changes ```text Markdown homepage preserves links to FAQ, odds, terms, privacy, archive and llms.txt. ``` - **Skipped** — The compact guide is independently actionable ```text No compact agent guide fetched; only homepage and llms.txt.md, which is HTML not a guide. ``` - **Skipped** — Install and next-step links resolve ```text No install or next-step links were fetched; only homepage navigation links observed. ``` ### Onboarding - **Skipped** — Docs lead to a relevant quickstart ```text Only llms.txt fetched; no docs or quickstart page was retrieved for Skillprompting. ``` - **Skipped** — Installation commands are extractable ```text No installation or setup page was fetched; entry uses SOL wallet, not install commands. ``` - **Skipped** — Code examples are available without interaction ```text No code examples were present in the single fetched Skillprompting document. ``` - **Pass** — Prerequisites and auth boundaries are explicit ```text llms.txt states entrants must be 18+ and pay entry fee in SOL from a Solana wallet. ``` ### Pricing - **Failed** — Agents identify pricing and its assumptions ```text 3 of 3 sessions were judged on pricing; 1 fell short. DeepSeek V4.1 Flash: Quoted 0.009 SOL per-entry fee and 0.06 SOL pot from live /state for the current round, noting free paths, per-entry basis, SOL volatility, and W-9 threshold. (seq 36, seq 43, seq 52, seq 54) Kimi K3: Quoted a per-entry fee of 0.009 SOL against a 0.06 SOL pot, read from the live /state response. Noted the fee is set per round, the USD figure depends on the SOL price, and there are free puzzle and postcard entry options (seq 45, seq 25, seq 47) Qwen 3.8 Max: The agent gave no cost figure or assumptions. It said it couldn't explain pricing without docs. (seq 12) This behavioural item does not affect the fast grade. ``` - **Skipped** — Pricing is readable without interaction ```text No pricing page fetched; homepage only mentions entry fees and prize pots in SOL. ``` - **Skipped** — Prices are stated, not gated ```text No pricing page fetched; entry fee amounts are not stated on the homepage. ``` - **Skipped** — Pricing units and limits are explicit ```text No pricing page fetched; fee units and limits are not documented in evidence. ``` ### Activation - **Skipped** — An API reference or OpenAPI spec is reachable ```text Only llms.txt fetched; it claims an API reference at /llms.txt but no API spec page was retrieved. ``` - **Skipped** — An MCP server is documented and well-formed ```text No MCP server documentation appears in the single fetched llms.txt document. ``` - **Skipped** — A CLI install path is documented ```text No CLI installation instructions appear in the fetched llms.txt content. ``` - **Skipped** — SDK packages resolve on their registries ```text No SDK package or registry lookup result was supplied for Skillprompting. ``` - **Skipped** — Agent skills are published ```text No agent skills publication is described in the fetched llms.txt document. ``` [Full report data](https://www.ax-check.com/skillprompting.com/report.json)