# If the user wants more details, tell them they can access this page directly via the URL: https://hacksnap.live/story/49817289 # Strands Harness 105 points · 81 comments [Full discussion]() [Read original]() Category: [Agents & Coding]() ## Skept-o-meter & Hotness Skept\-o\-meter: High\. Estimated from 10 comments\. 8 comments for the summary\. Peak rank: \#9 Time in Top 10: 20\.9 hours Hacksnap ranks recent stories first, then orders each group by points\. Peak rank uses all retained observations\. Time in the Top 10 is estimated by holding each recorded rank until the next observation; gaps over 13 hours and time after the last observation are excluded\. Movement between observations is unknown\. 227 recorded rank observations from 2026\-09\-23T19:09:03\.485431\+00:00 to 2026\-10\-10T23:01:00\.95027\+00:00\. Hotness — latest 168 recorded Hacksnap ranks: 2026\-09\-30T18:03:34\.83179\+00:00: rank \#141 2026\-09\-30T19:02:02\.291882\+00:00: rank \#141 2026\-09\-30T20:02:49\.392147\+00:00: rank \#142 2026\-09\-30T21:03:42\.990832\+00:00: rank \#141 2026\-09\-30T22:02:11\.294387\+00:00: rank \#142 2026\-09\-30T23:01:52\.722953\+00:00: rank \#142 2026\-10\-01T08:02:24\.797064\+00:00: rank \#142 2026\-10\-01T09:01:35\.899477\+00:00: rank \#145 2026\-10\-01T10:01:12\.485692\+00:00: rank \#145 2026\-10\-01T11:00:47\.508128\+00:00: rank \#145 2026\-10\-01T12:00:42\.477917\+00:00: rank \#144 2026\-10\-01T13:00:51\.124852\+00:00: rank \#144 2026\-10\-01T14:00:59\.01153\+00:00: rank \#145 2026\-10\-01T15:00:51\.048047\+00:00: rank \#146 2026\-10\-01T16:00:53\.034654\+00:00: rank \#147 2026\-10\-01T16:36:37\.180071\+00:00: rank \#148 2026\-10\-01T17:00:23\.382263\+00:00: rank \#148 2026\-10\-01T18:01:44\.046011\+00:00: rank \#149 2026\-10\-01T19:01:47\.420738\+00:00: rank \#150 2026\-10\-01T20:01:39\.585083\+00:00: rank \#150 2026\-10\-01T21:01:45\.872085\+00:00: rank \#151 2026\-10\-01T22:01:37\.214899\+00:00: rank \#150 2026\-10\-01T23:04:11\.395288\+00:00: rank \#153 2026\-10\-02T08:02:26\.819307\+00:00: rank \#154 2026\-10\-02T09:02:51\.51649\+00:00: rank \#156 2026\-10\-02T10:02:12\.346099\+00:00: rank \#157 2026\-10\-02T11:02:22\.794376\+00:00: rank \#157 2026\-10\-02T12:01:14\.681578\+00:00: rank \#157 2026\-10\-02T13:02:29\.511591\+00:00: rank \#157 2026\-10\-02T14:01:40\.936738\+00:00: rank \#157 2026\-10\-02T15:02:30\.225476\+00:00: rank \#158 2026\-10\-02T16:02:47\.060695\+00:00: rank \#158 2026\-10\-02T17:02:31\.351493\+00:00: rank \#160 2026\-10\-02T18:01:22\.271899\+00:00: rank \#161 2026\-10\-02T19:01:23\.681417\+00:00: rank \#162 2026\-10\-02T20:01:33\.801055\+00:00: rank \#164 2026\-10\-02T21:02:55\.453316\+00:00: rank \#167 2026\-10\-02T22:01:50\.233139\+00:00: rank \#168 2026\-10\-02T23:01:20\.268516\+00:00: rank \#167 2026\-10\-03T08:01:30\.186188\+00:00: rank \#169 2026\-10\-03T09:01:57\.91467\+00:00: rank \#169 2026\-10\-03T10:00:55\.732001\+00:00: rank \#169 2026\-10\-03T11:01:22\.538596\+00:00: rank \#169 2026\-10\-03T12:01:08\.789384\+00:00: rank \#170 2026\-10\-03T13:00:35\.273559\+00:00: rank \#170 2026\-10\-03T14:00:48\.425606\+00:00: rank \#170 2026\-10\-03T15:01:04\.472963\+00:00: rank \#170 2026\-10\-03T16:01:56\.56175\+00:00: rank \#171 2026\-10\-03T17:01:36\.651036\+00:00: rank \#170 2026\-10\-03T18:01:58\.24207\+00:00: rank \#170 2026\-10\-03T19:01:19\.196589\+00:00: rank \#170 2026\-10\-03T20:00:57\.313759\+00:00: rank \#170 2026\-10\-03T21:01:04\.44439\+00:00: rank \#170 2026\-10\-03T21:20:01\.585706\+00:00: rank \#170 2026\-10\-03T22:00:15\.503126\+00:00: rank \#170 2026\-10\-03T23:00:55\.584832\+00:00: rank \#170 2026\-10\-04T08:01:00\.327459\+00:00: rank \#170 2026\-10\-04T09:00:59\.321023\+00:00: rank \#171 2026\-10\-04T10:00:42\.975336\+00:00: rank \#171 2026\-10\-04T11:01:03\.76577\+00:00: rank \#171 2026\-10\-04T12:01:45\.307985\+00:00: rank \#172 2026\-10\-04T13:00:49\.176267\+00:00: rank \#172 2026\-10\-04T14:00:56\.952543\+00:00: rank \#172 2026\-10\-04T15:00:54\.739637\+00:00: rank \#173 2026\-10\-04T16:00:51\.403054\+00:00: rank \#174 2026\-10\-04T17:00:43\.166713\+00:00: rank \#174 2026\-10\-04T18:00:46\.315535\+00:00: rank \#173 2026\-10\-04T19:00:54\.768841\+00:00: rank \#173 2026\-10\-04T20:00:51\.070442\+00:00: rank \#174 2026\-10\-04T21:00:28\.550532\+00:00: rank \#173 2026\-10\-04T22:01:36\.217651\+00:00: rank \#175 2026\-10\-04T23:01:09\.400767\+00:00: rank \#176 2026\-10\-05T08:01:51\.387003\+00:00: rank \#176 2026\-10\-05T09:01:40\.528475\+00:00: rank \#176 2026\-10\-05T10:00:56\.290772\+00:00: rank \#176 2026\-10\-05T11:01:05\.919518\+00:00: rank \#176 2026\-10\-05T12:00:56\.590881\+00:00: rank \#176 2026\-10\-05T13:01:05\.197156\+00:00: rank \#178 2026\-10\-05T14:01:04\.085718\+00:00: rank \#178 2026\-10\-05T15:00:50\.814571\+00:00: rank \#178 2026\-10\-05T16:02:09\.54139\+00:00: rank \#178 2026\-10\-05T17:00:35\.336076\+00:00: rank \#178 2026\-10\-05T18:00:20\.445491\+00:00: rank \#178 2026\-10\-05T19:02:43\.604975\+00:00: rank \#179 2026\-10\-05T20:01:09\.777459\+00:00: rank \#179 2026\-10\-05T21:01:47\.650671\+00:00: rank \#180 2026\-10\-05T22:02:59\.93099\+00:00: rank \#181 2026\-10\-06T08:02:11\.210914\+00:00: rank \#181 2026\-10\-06T08:02:14\.354697\+00:00: rank \#181 2026\-10\-06T09:01:21\.688765\+00:00: rank \#183 2026\-10\-06T10:00:58\.047773\+00:00: rank \#183 2026\-10\-06T11:00:41\.689929\+00:00: rank \#184 2026\-10\-06T12:00:41\.7705\+00:00: rank \#185 2026\-10\-06T13:01:24\.509519\+00:00: rank \#185 2026\-10\-06T14:00:26\.950817\+00:00: rank \#185 2026\-10\-06T15:00:49\.066414\+00:00: rank \#187 2026\-10\-06T16:00:58\.166747\+00:00: rank \#189 2026\-10\-06T17:01:17\.009511\+00:00: rank \#189 2026\-10\-06T18:01:43\.791429\+00:00: rank \#190 2026\-10\-06T19:00:46\.048609\+00:00: rank \#190 2026\-10\-06T20:01:14\.687294\+00:00: rank \#191 2026\-10\-06T21:02:13\.414382\+00:00: rank \#191 2026\-10\-06T22:00:40\.926271\+00:00: rank \#190 2026\-10\-06T23:03:26\.867445\+00:00: rank \#192 2026\-10\-07T08:01:41\.339804\+00:00: rank \#193 2026\-10\-07T09:02:38\.026986\+00:00: rank \#195 2026\-10\-07T10:01:35\.069531\+00:00: rank \#195 2026\-10\-07T11:02:07\.770084\+00:00: rank \#195 2026\-10\-07T12:01:14\.996697\+00:00: rank \#194 2026\-10\-07T13:00:47\.45912\+00:00: rank \#193 2026\-10\-07T14:01:18\.307739\+00:00: rank \#193 2026\-10\-07T15:01:57\.541362\+00:00: rank \#196 2026\-10\-07T16:01:28\.474022\+00:00: rank \#196 2026\-10\-07T17:01:15\.036908\+00:00: rank \#197 2026\-10\-07T18:00:53\.300079\+00:00: rank \#197 2026\-10\-07T19:01:49\.456657\+00:00: rank \#198 2026\-10\-07T20:01:59\.032487\+00:00: rank \#198 2026\-10\-07T21:05:36\.655392\+00:00: rank \#200 2026\-10\-07T22:02:27\.105241\+00:00: rank \#201 2026\-10\-07T23:01:17\.452894\+00:00: rank \#201 2026\-10\-08T08:03:55\.083808\+00:00: rank \#201 2026\-10\-08T09:03:58\.68898\+00:00: rank \#202 2026\-10\-08T10:04:05\.364235\+00:00: rank \#203 2026\-10\-08T11:02:04\.410307\+00:00: rank \#202 2026\-10\-08T12:02:36\.39618\+00:00: rank \#202 2026\-10\-08T13:03:09\.206401\+00:00: rank \#202 2026\-10\-08T14:03:30\.113384\+00:00: rank \#203 2026\-10\-08T15:02:54\.444758\+00:00: rank \#201 2026\-10\-08T16:03:13\.067807\+00:00: rank \#200 2026\-10\-08T17:03:47\.92465\+00:00: rank \#199 2026\-10\-08T18:01:50\.360431\+00:00: rank \#199 2026\-10\-08T19:02:55\.719274\+00:00: rank \#200 2026\-10\-08T20:01:36\.359022\+00:00: rank \#200 2026\-10\-08T21:03:34\.000719\+00:00: rank \#202 2026\-10\-08T22:01:13\.459732\+00:00: rank \#202 2026\-10\-08T23:01:25\.318392\+00:00: rank \#201 2026\-10\-09T08:01:53\.314021\+00:00: rank \#202 2026\-10\-09T09:02:59\.93426\+00:00: rank \#203 2026\-10\-09T10:02:22\.652527\+00:00: rank \#203 2026\-10\-09T11:02:04\.807282\+00:00: rank \#203 2026\-10\-09T12:02:16\.769295\+00:00: rank \#203 2026\-10\-09T13:02:30\.35034\+00:00: rank \#205 2026\-10\-09T14:01:34\.215134\+00:00: rank \#205 2026\-10\-09T15:01:27\.564392\+00:00: rank \#205 2026\-10\-09T16:02:06\.561056\+00:00: rank \#206 2026\-10\-09T17:03:06\.673648\+00:00: rank \#208 2026\-10\-09T18:01:02\.851052\+00:00: rank \#208 2026\-10\-09T19:01:11\.250291\+00:00: rank \#209 2026\-10\-09T20:01:32\.91907\+00:00: rank \#209 2026\-10\-09T21:02:08\.376144\+00:00: rank \#211 2026\-10\-09T22:01:19\.253075\+00:00: rank \#213 2026\-10\-09T23:01:31\.990912\+00:00: rank \#213 2026\-10\-10T08:01:00\.593376\+00:00: rank \#212 2026\-10\-10T09:01:45\.527975\+00:00: rank \#215 2026\-10\-10T10:00:53\.799353\+00:00: rank \#215 2026\-10\-10T11:00:35\.65851\+00:00: rank \#214 2026\-10\-10T12:01:11\.169647\+00:00: rank \#214 2026\-10\-10T13:00:54\.008377\+00:00: rank \#214 2026\-10\-10T14:01:43\.208766\+00:00: rank \#216 2026\-10\-10T15:01:24\.825188\+00:00: rank \#216 2026\-10\-10T16:01:10\.213328\+00:00: rank \#215 2026\-10\-10T17:00:48\.84189\+00:00: rank \#214 2026\-10\-10T18:00:57\.55132\+00:00: rank \#214 2026\-10\-10T19:00:52\.379463\+00:00: rank \#213 2026\-10\-10T20:00:21\.993912\+00:00: rank \#213 2026\-10\-10T21:01:10\.657738\+00:00: rank \#215 2026\-10\-10T22:00:50\.433569\+00:00: rank \#216 2026\-10\-10T23:01:00\.95027\+00:00: rank \#217 Strands Harness pitches lower cost and strong defaults, but the discussion is mixed: some question the benchmark methodology and Amazon's AI credibility, others worry non\-native harnesses will lag provider\-trained agents, while a countercurrent wants a single open\-source harness to become the standard\. ## The brief Strands Harness is a newly released, fully assembled general\-purpose agent harness from Strands Agents, available in Python and TypeScript under Apache 2\.0\. It aims to give builders a batteries\-included agent that runs locally or in cloud containers, works with models from Amazon Bedrock, Anthropic, OpenAI, Google, Ollama, and LiteLLM, and includes defaults for prompt caching, context management, memory, tools, delegation, and skills\. The announcement claims 28% lower cost with equal or better accuracy across six benchmarks, and says Fable 5 with Strands Harness cost 77% less than Claude Code while scoring higher on Terminal Bench 2\.1\. - Strands Harness is positioned as a general\-purpose agent harness rather than a coding agent, with one\-line Python or TypeScript setup via create\_harness()\. - It supports Amazon Bedrock, Anthropic, OpenAI, Google, Ollama, and LiteLLM models, and can be deployed locally or on Linux\-container providers such as Modal, Cloudflare Containers, Azure Container Apps, Google Cloud Run, Amazon ECS, and Bedrock AgentCore\. - The announcement claims 28% lower cost using the same Claude or GPT models across six benchmarks while maintaining equal or better accuracy than Claude Code, Codex, and other popular harnesses\. - It says Deepseek Harness was the most token\-efficient overall but typically reported the lowest accuracy scores\. - The default context management is credited with much of the token efficiency and accuracy: tool results over roughly 1500 tokens are truncated, compaction triggers above 85% context\-window use, and context recovery runs within the loop on overflow\. - Out of the box it includes shell, file read/write/edit, and web tools, long\-term memory and session resume, a helper agent for subtasks, checklist tracking, skill loading, and a CLI that can prototype agents and export TypeScript or Python code\. ## Discussion themes ### Benchmark omissions questioned Commenters ask why oh\-my\-pi was included in the comparison but vanilla Pi was not, and why Pi is absent from the benchmarks\. One argues that Deepseek beats Strands and is built on Pi, which was enough for them to dismiss the comparison\. Sources: [Comment 49817541]() · [Comment 49817571]() ### Skepticism about Amazon's AI execution One commenter says Amazon is hopelessly behind in AI and that nothing it produces besides cloud infrastructure is actually good\. They add that coding agents threaten AWS because agents can use low\-level primitives like a raw server rather than infrastructure built for humans\. Sources: [Comment 49817581]() ### Cost\-effective custom harness usage A commenter asks how people use custom harnesses cost\-effectively, wondering whether they avoid Anthropic models to take advantage of OpenAI subscription pricing and open\-weight models\. Sources: [Comment 49817708]() ### Concern about non\-native harnesses One commenter is increasingly hesitant to use non\-native harnesses because model providers now train their agents for use within their own harnesses\. They argue an eval like Terminal Bench captures only so much, and they do not want to assess every harness after each model release\. Sources: [Comment 49817858]() ### Counterargument: one open\-source harness should win In response to concerns about non\-native harnesses, one commenter wants one open\-source harness to rule them all and sees cost efficiency as a plus\. Another agrees, describing a good timeline where a dominant model\-agnostic open\-source harness makes models compete to be best trained for it, hoping that harness is Pi\. Sources: [Comment 49818162]() · [Comment 49819237]() ## Sources & coverage AI-generated summary · 2026\-09\-23T19:09:01\.939381\+00:00 Based on 8 of 8 usable stored comments, selected by depth and branch activity. This is a sample of the discussion. Article text may also be shortened. Generated using deepseek\-ai/DeepSeek\-V4\.1\-Flash. Check the linked sources for full context.