{"type":"mcp-server-v1","name":"Valory Mech Social (Robinhood, service 3)","description":"The mech executes AI tasks requested on-chain and delivers the results to the requester.","version":"1.0.0","endpoint":"https://7ab7034285fb57610053125c9e71f9f3738eb1f7-4663.mech.valory.xyz","capabilities":{"tools":{"listChanged":false}},"metadata":{"network":"robinhood-chain","serviceId":"3","sourceToolMetadataCid":"f0170122095a3ca8204796805fdca9f5e5ed4403c138ade0d99897a6f52c66fbc5071d006","generatedAt":"2026-11-19T07:49:48.158Z","marketplaceAddress":"eip155:4663:0xa45E64d13A30a51b91ae0eb182e88a40e9b18eD8","mechAddress":"eip155:4663:0x7ab7034285fb57610053125c9e71f9f3738eb1f7"},"tools":[{"name":"token_social_sentiment","description":"Measures social sentiment about the ONE token named in the request, from X posts and news headlines. Input: JSON {\"symbol\", \"address\", \"chain\", \"window_seconds\"} (symbol or EVM address required; window defaults to 24h, 3600..604740, or \"past week\" style periods in free text) or free text such as \"How is sentiment on $PEPE today?\". No token discovery or ranking, one token per request. Output: JSON with sentiment (-1 to 1, null below 5 on-topic items), breakdown counts, mentions, mentions_trend, posts_analyzed, top_posts, headlines, reasoning, warnings, degraded_sources and error.","inputSchema":{"type":"string","description":"One token per request, named by the caller: a JSON object {\"symbol\", \"address\", \"chain\", \"window_seconds\"} with symbol or address required, or free text such as \"How is sentiment on $PEPE today?\". An EVM address gives the most exact answer. window_seconds defaults to 86400 (24 h) and is clamped to 3600..604740, so ask for 3600 for \"right now\" and 604740 for a week; in free text an explicit period such as \"over the past week\" or \"last 3 days\" sets it. The tool does not discover or rank tokens, does not answer questions about several tokens, and returns no list of scam or rug-pull warnings: it measures the sentiment of the token named in the request over the window."},"outputSchema":{"type":"object","description":"A JSON object containing the social sentiment for the token","properties":{"requestId":{"type":"integer","description":"Unique identifier for the request"},"result":{"type":"string","description":"Result in JSON format as a string, always with every key. sentiment: (bullish - bearish) / on-topic sample items, -1 to 1, null below 5 on-topic items. sentiment_interval: {\"low\": x, \"high\": y}, a 95% confidence interval for the mean of this sample: when two readings' intervals do not overlap the difference is likely real, and overlap does not show they agree; not the range a rerun would land in; null when sentiment is null. breakdown: number of on-topic sample items (posts plus headlines) per class, not mentions. mentions: all X posts matching the search in the window, spam included (bot templates are excluded in the search itself). mentions_trend: {\"recent\": n, \"previous\": m}, the matching posts in the newer and the older half of the window; null when the count is unavailable, when the window is under 4 hours, and when the hourly buckets cover too little of it (then degraded_sources carries \"x_trend\"). posts_analyzed: on-topic X posts in the sample (at most 40). warnings: list of {type, message}, type symbol_mismatch, chain_mismatch, symbol_unverified, address_not_listed or busier_token_same_ticker (a thin or unlisted address; the message names the token with the most DEX volume for the ticker). degraded_sources: sources that failed while a result was still produced (x, x_partial, x_counts, x_trend, news, dexscreener). error: null or {type, message}, type invalid_input, source_unavailable, llm_error or internal.","example":"{\n  \"token\": \"ABC\",\n  \"address\": \"0x1111111111111111111111111111111111111111\",\n  \"chain\": \"base\",\n  \"window_seconds\": 86400,\n  \"sentiment\": 0.5,\n  \"sentiment_interval\": {\"low\": 0.07, \"high\": 0.93},\n  \"mentions\": 1250,\n  \"mentions_trend\": {\"recent\": 800, \"previous\": 450},\n  \"posts_analyzed\": 9,\n  \"breakdown\": {\"bullish\": 6, \"neutral\": 3, \"bearish\": 1},\n  \"reasoning\": \"Most on-topic posts react positively to a new exchange listing.\",\n  \"top_posts\": [\n    \"https://x.com/i/web/status/1000000000000000001\",\n    \"https://x.com/i/web/status/1000000000000000002\",\n    \"https://x.com/i/web/status/1000000000000000003\",\n    \"https://x.com/i/web/status/1000000000000000004\",\n    \"https://x.com/i/web/status/1000000000000000005\"\n  ],\n  \"headlines\": [{\"title\": \"ABC lists on a new exchange\", \"url\": \"https://example.com/abc\"}],\n  \"warnings\": [],\n  \"degraded_sources\": [],\n  \"error\": null\n}"},"prompt":{"type":"string","description":"The prompt passed to the tool."}},"required":["requestId","result","prompt"]}}]}