[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M737H9YMHNGSPFYJSCMJBCBS":3,"skill-same:01M737H9YMHNGSPFYJSCMJBCBS":67},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":40,"commitSha":41,"skillDirectory":42,"forksCount":43,"repositoryHtmlUrl":44,"repositoryHomepageUrl":6,"ownerHtmlUrl":45,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":46,"install":47,"resources":60},"01M737H9YMHNGSPFYJSCMJBCBS","actual-setup",null,"Set up Actual Computer (actual.inc) inference in Hermes.","ashneil12\u002Fvanilla-hermes-agent-canary","ashneil12","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F49880658?v=4","01M7378BQ9W161EPENEWBYAQNN","01M7378BRDJG8SFS93FG4V6BWY",1,"MIT","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Foptional-skills\u002Fdevops\u002Factual-setup\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: actual-setup\ndescription: Set up Actual Computer (actual.inc) inference in Hermes.\nversion: 2.0.0\nauthor: shl0ms + Hermes Agent\nlicense: MIT\nplatforms: [linux, macos, windows]\nmetadata:\n  hermes:\n    tags: [actual, actual-inc, provider, local-inference, relay, gguf, setup]\n    category: devops\n---\n\n# Actual Computer Setup Skill\n\nSets up [actual.inc](https:\u002F\u002Factual.inc) (Actual Computer) as a Hermes inference\nprovider. Actual turns the user's own hardware into a private inference cluster\nand exposes an OpenAI-compatible API two ways: a hosted end-to-end-encrypted\nrelay at `https:\u002F\u002Fapi.actual.inc` (authenticated with an `ac_` key), and a local\non-device daemon at `http:\u002F\u002F127.0.0.1:8080` (no auth on loopback). This skill\ndoes not install the Actual daemon for the user — device authorization requires\na human in a browser.\n\n## When to Use\n\n- User wants to add actual.inc as an inference provider (cloud relay or local).\n- User has an `ac_` key and wants Hermes routed through their Actual cluster.\n- User wants fully-local, on-device inference via the Actual daemon.\n- Troubleshooting: Actual requests failing with cryptic 400s or empty streams.\n\n## Prerequisites\n\n- Hermes has **first-class `actual` provider support** (provider id `actual`,\n  aliases `actual-computer`, `actualcomputer`, `aci`). Do NOT configure Actual\n  as a `custom_providers` \u002F `providers.actual.*` entry on current Hermes — the\n  built-in provider owns the name and handles base-url normalization, the\n  Responses transport, and local no-auth automatically.\n- Relay mode: an Actual account and an `ac_` inference key from\n  https:\u002F\u002Factual.inc\u002Fuser\u002Fkeys.\n- Local mode: the user has installed the daemon\n  (`curl -fsSL \"https:\u002F\u002Factual.inc\u002Finstall\" | bash`) and completed device\n  authorization by running `actual` once and opening the printed\n  `https:\u002F\u002Factual.inc\u002Fdevice?code=...` URL in a browser. Relay that URL to the\n  user and WAIT — never invent an email or authorize on their behalf. Codes\n  expire in 5 minutes; re-run `actual` for a fresh one.\n\n## How to Run\n\n### Relay \u002F API mode\n\n1. Put the key in `.env` (secrets only — never config.yaml):\n   append `ACTUAL_API_KEY=ac_...` to `~\u002F.hermes\u002F.env`.\n2. Verify the key and discover models with `terminal`:\n   ```bash\n   curl -s https:\u002F\u002Fapi.actual.inc\u002Fv1\u002Fmodels -H \"Authorization: Bearer $ACTUAL_API_KEY\"\n   ```\n3. Select provider + model:\n   ```bash\n   hermes config set model.provider actual\n   hermes config set model.default \"MODEL_ID_FROM_DISCOVERY\"\n   ```\n4. Verify end-to-end:\n   ```bash\n   hermes chat -Q -q \"Reply with exactly: ACTUAL_OK\" --provider actual -m MODEL_ID\n   ```\n\n### Local mode\n\n1. Human has installed + authorized the daemon (see Prerequisites).\n2. Download and load a model (scriptable once authorized):\n   ```bash\n   actual models search \"qwen2.5 0.5b instruct gguf\" --limit 8 --no-prompt\n   # Downloads REQUIRE an explicit quantization (409 ambiguous_model_download otherwise):\n   actual models download \"Qwen\u002FQwen2.5-0.5B-Instruct-GGUF\u002FQ4_K_M\"\n   actual models list        # note the INSTALLED name (differs from download id)\n   actual models load \"qwen2.5-0.5b-instruct-q4_k_m\"   # load by installed name\n   ```\n3. Point Hermes at the daemon. `ACTUAL_BASE_URL` with a loopback host flips the\n   built-in provider into local no-auth mode automatically — no key needed:\n   append `ACTUAL_BASE_URL=http:\u002F\u002F127.0.0.1:8080` to `~\u002F.hermes\u002F.env`, then:\n   ```bash\n   hermes config set model.provider actual\n   hermes config set model.default \"INSTALLED_MODEL_NAME\"\n   ```\n4. Verify (reduced toolset — see context-window pitfall below):\n   ```bash\n   hermes chat -Q -q \"Reply with exactly: LOCAL_OK\" --provider actual -m INSTALLED_NAME -t file,web\n   ```\n\n## Quick Reference\n\n| Thing | Value |\n|---|---|\n| Hosted relay | `https:\u002F\u002Fapi.actual.inc\u002Fv1` (normalized from bare host automatically) |\n| Local daemon | `http:\u002F\u002F127.0.0.1:8080\u002Fv1` (no auth on loopback) |\n| Key env var | `ACTUAL_API_KEY` (`ac_...`) |\n| Base URL env var | `ACTUAL_BASE_URL` (loopback host ⇒ local no-auth mode) |\n| Provider id \u002F aliases | `actual` \u002F `actual-computer`, `actualcomputer`, `aci` |\n| Transport | Responses API (`codex_responses`) — built-in, do not override |\n| Cluster pinning | `X-Cluster-ID` header via `providers.actual.extra_headers` in config.yaml |\n| Model size guide | 0.5B Q4_K_M ~470MB (toy), 7-8B Q4_K_M ~4.5GB (daily driver), 32B ~20GB |\n\n## Pitfalls\n\n1. **reasoning_effort trap (handled by Hermes since the first-class provider).**\n   Actual's SGLang\u002FvLLM backends accept only `none\u002Flow\u002Fmedium\u002Fhigh\u002Fmax`;\n   `xhigh`\u002F`ultra` used to fail with a cryptic\n   `Expecting value: line 1 column 1 (char 0)` (a wrapped HTTP 400). The\n   built-in provider clamps `xhigh→high` and `ultra→max` on the wire. If a\n   request still 400s this way on an old Hermes, set a per-model cap:\n   `agent.reasoning_overrides.\u003Cmodel>: high` in config.yaml.\n2. **Context-window overflow on small local models.** Hermes' default toolset\n   is ~26k tokens of schemas plus a ~9k-token system prompt. A model loaded\n   with a 32k context overflows before the first turn, and llama.cpp-family\n   servers emit a bare `data: [DONE]` — Hermes reports\n   `Provider returned an empty stream with no finish_reason`. This is NOT an\n   SSE bug. Fixes: restrict tools (`-t file,web`), load the model with a\n   larger `n_ctx`, or pick a >=64k-context model for the full toolset.\n   Upstream tracking: #51448 (do not file new issues; add evidence there).\n   Related but distinct: #65631 (HTTP-200 SSE carrying a 400), #56516\n   (reasoning-only streams).\n3. **Download ids vs installed names.** `actual models download` takes\n   `repo\u002FQUANT` and 409s without an explicit quantization;\n   `actual models load` takes the INSTALLED name from `actual models list`.\n4. **Reasoning models returning empty content.** GLM\u002FQwen reasoning variants\n   emit thinking in a separate `reasoning` field and can burn a small\n   `max_tokens` entirely on reasoning. Give generous max_tokens before\n   assuming failure.\n5. **Do not create a custom provider named `actual`.** Older setup guides\n   (pre first-class support) wrote `providers.actual.*` config blocks. On\n   current Hermes the built-in provider wins the name; stale custom blocks\n   are ignored or conflict. Remove them and use the env vars + model.provider\n   flow above.\n\n## Verification\n\n```bash\n# Relay:\nhermes chat -Q -q \"Reply with exactly: ACTUAL_OK\" --provider actual -m MODEL\n# Local (small model — reduced toolset):\nhermes chat -Q -q \"Reply with exactly: LOCAL_OK\" --provider actual -m MODEL -t file,web\n# Provider status (local no-auth shows key_source=local-offline):\nhermes status\n```\n\nFor other OpenAI-compatible clients (e.g. OpenCode), see\n`references\u002Fopencode.md`.\n","# Actual Computer Setup Skill\n\nSets up [actual.inc](https:\u002F\u002Factual.inc) (Actual Computer) as a Hermes inference\nprovider. Actual turns the user's own hardware into a private inference cluster\nand exposes an OpenAI-compatible API two ways: a hosted end-to-end-encrypted\nrelay at `https:\u002F\u002Fapi.actual.inc` (authenticated with an `ac_` key), and a local\non-device daemon at `http:\u002F\u002F127.0.0.1:8080` (no auth on loopback). This skill\ndoes not install the Actual daemon for the user — device authorization requires\na human in a browser.\n\n## When to Use\n\n- User wants to add actual.inc as an inference provider (cloud relay or local).\n- User has an `ac_` key and wants Hermes routed through their Actual cluster.\n- User wants fully-local, on-device inference via the Actual daemon.\n- Troubleshooting: Actual requests failing with cryptic 400s or empty streams.\n\n## Prerequisites\n\n- Hermes has **first-class `actual` provider support** (provider id `actual`,\n  aliases `actual-computer`, `actualcomputer`, `aci`). Do NOT configure Actual\n  as a `custom_providers` \u002F `providers.actual.*` entry on current Hermes — the\n  built-in provider owns the name and handles base-url normalization, the\n  Responses transport, and local no-auth automatically.\n- Relay mode: an Actual account and an `ac_` inference key from\n  https:\u002F\u002Factual.inc\u002Fuser\u002Fkeys.\n- Local mode: the user has installed the daemon\n  (`curl -fsSL \"https:\u002F\u002Factual.inc\u002Finstall\" | bash`) and completed device\n  authorization by running `actual` once and opening the printed\n  `https:\u002F\u002Factual.inc\u002Fdevice?code=...` URL in a browser. Relay that URL to the\n  user and WAIT — never invent an email or authorize on their behalf. Codes\n  expire in 5 minutes; re-run `actual` for a fresh one.\n\n## How to Run\n\n### Relay \u002F API mode\n\n1. Put the key in `.env` (secrets only — never config.yaml):\n   append `ACTUAL_API_KEY=ac_...` to `~\u002F.hermes\u002F.env`.\n2. Verify the key and discover models with `terminal`:\n   ```bash\n   curl -s https:\u002F\u002Fapi.actual.inc\u002Fv1\u002Fmodels -H \"Authorization: Bearer $ACTUAL_API_KEY\"\n   ```\n3. Select provider + model:\n   ```bash\n   hermes config set model.provider actual\n   hermes config set model.default \"MODEL_ID_FROM_DISCOVERY\"\n   ```\n4. Verify end-to-end:\n   ```bash\n   hermes chat -Q -q \"Reply with exactly: ACTUAL_OK\" --provider actual -m MODEL_ID\n   ```\n\n### Local mode\n\n1. Human has installed + authorized the daemon (see Prerequisites).\n2. Download and load a model (scriptable once authorized):\n   ```bash\n   actual models search \"qwen2.5 0.5b instruct gguf\" --limit 8 --no-prompt\n   # Downloads REQUIRE an explicit quantization (409 ambiguous_model_download otherwise):\n   actual models download \"Qwen\u002FQwen2.5-0.5B-Instruct-GGUF\u002FQ4_K_M\"\n   actual models list        # note the INSTALLED name (differs from download id)\n   actual models load \"qwen2.5-0.5b-instruct-q4_k_m\"   # load by installed name\n   ```\n3. Point Hermes at the daemon. `ACTUAL_BASE_URL` with a loopback host flips the\n   built-in provider into local no-auth mode automatically — no key needed:\n   append `ACTUAL_BASE_URL=http:\u002F\u002F127.0.0.1:8080` to `~\u002F.hermes\u002F.env`, then:\n   ```bash\n   hermes config set model.provider actual\n   hermes config set model.default \"INSTALLED_MODEL_NAME\"\n   ```\n4. Verify (reduced toolset — see context-window pitfall below):\n   ```bash\n   hermes chat -Q -q \"Reply with exactly: LOCAL_OK\" --provider actual -m INSTALLED_NAME -t file,web\n   ```\n\n## Quick Reference\n\n| Thing | Value |\n|---|---|\n| Hosted relay | `https:\u002F\u002Fapi.actual.inc\u002Fv1` (normalized from bare host automatically) |\n| Local daemon | `http:\u002F\u002F127.0.0.1:8080\u002Fv1` (no auth on loopback) |\n| Key env var | `ACTUAL_API_KEY` (`ac_...`) |\n| Base URL env var | `ACTUAL_BASE_URL` (loopback host ⇒ local no-auth mode) |\n| Provider id \u002F aliases | `actual` \u002F `actual-computer`, `actualcomputer`, `aci` |\n| Transport | Responses API (`codex_responses`) — built-in, do not override |\n| Cluster pinning | `X-Cluster-ID` header via `providers.actual.extra_headers` in config.yaml |\n| Model size guide | 0.5B Q4_K_M ~470MB (toy), 7-8B Q4_K_M ~4.5GB (daily driver), 32B ~20GB |\n\n## Pitfalls\n\n1. **reasoning_effort trap (handled by Hermes since the first-class provider).**\n   Actual's SGLang\u002FvLLM backends accept only `none\u002Flow\u002Fmedium\u002Fhigh\u002Fmax`;\n   `xhigh`\u002F`ultra` used to fail with a cryptic\n   `Expecting value: line 1 column 1 (char 0)` (a wrapped HTTP 400). The\n   built-in provider clamps `xhigh→high` and `ultra→max` on the wire. If a\n   request still 400s this way on an old Hermes, set a per-model cap:\n   `agent.reasoning_overrides.\u003Cmodel>: high` in config.yaml.\n2. **Context-window overflow on small local models.** Hermes' default toolset\n   is ~26k tokens of schemas plus a ~9k-token system prompt. A model loaded\n   with a 32k context overflows before the first turn, and llama.cpp-family\n   servers emit a bare `data: [DONE]` — Hermes reports\n   `Provider returned an empty stream with no finish_reason`. This is NOT an\n   SSE bug. Fixes: restrict tools (`-t file,web`), load the model with a\n   larger `n_ctx`, or pick a >=64k-context model for the full toolset.\n   Upstream tracking: #51448 (do not file new issues; add evidence there).\n   Related but distinct: #65631 (HTTP-200 SSE carrying a 400), #56516\n   (reasoning-only streams).\n3. **Download ids vs installed names.** `actual models download` takes\n   `repo\u002FQUANT` and 409s without an explicit quantization;\n   `actual models load` takes the INSTALLED name from `actual models list`.\n4. **Reasoning models returning empty content.** GLM\u002FQwen reasoning variants\n   emit thinking in a separate `reasoning` field and can burn a small\n   `max_tokens` entirely on reasoning. Give generous max_tokens before\n   assuming failure.\n5. **Do not create a custom provider named `actual`.** Older setup guides\n   (pre first-class support) wrote `providers.actual.*` config blocks. On\n   current Hermes the built-in provider wins the name; stale custom blocks\n   are ignored or conflict. Remove them and use the env vars + model.provider\n   flow above.\n\n## Verification\n\n```bash\n# Relay:\nhermes chat -Q -q \"Reply with exactly: ACTUAL_OK\" --provider actual -m MODEL\n# Local (small model — reduced toolset):\nhermes chat -Q -q \"Reply with exactly: LOCAL_OK\" --provider actual -m MODEL -t file,web\n# Provider status (local no-auth shows key_source=local-offline):\nhermes status\n```\n\nFor other OpenAI-compatible clients (e.g. OpenCode), see\n`references\u002Fopencode.md`.\n",{"name":5,"author":23,"license":14,"version":24,"metadata":25,"platforms":36,"description":7},"shl0ms + Hermes Agent","2.0.0",{"hermes":26},{"tags":27,"category":35},[28,29,30,31,32,33,34],"actual","actual-inc","provider","local-inference","relay","gguf","setup","devops",[37,38,39],"linux","macos","windows","main","0905b0264ce5a54908bedc9637e50251543db132","optional-skills\u002Fdevops\u002Factual-setup",0,"https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary","https:\u002F\u002Fgithub.com\u002Fashneil12","2026-11-04 08:25:00",{"source":8,"skillName":5,"commands":48,"prompt":58,"zipAvailable":59},[49,52,55],{"tool":50,"command":51},"npx","npx skills add ashneil12\u002Fvanilla-hermes-agent-canary --skill actual-setup",{"tool":53,"command":54},"bunx","bunx skills add ashneil12\u002Fvanilla-hermes-agent-canary --skill actual-setup",{"tool":56,"command":57},"pnpm","pnpm dlx skills add ashneil12\u002Fvanilla-hermes-agent-canary --skill actual-setup","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Foptional-skills\u002Fdevops\u002Factual-setup\u002FSKILL.md",true,[61],{"path":62,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":65,"sourceHtmlUrl":66},"references\u002Fopencode.md","REFERENCE","参考文档",3720,"https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Foptional-skills\u002Fdevops\u002Factual-setup\u002Freferences\u002Fopencode.md",[68,75,82,89,96,103],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M7380YR13A23353XT9NS6EGR","blocked-page-recovery","Use when a fetch fails: 403\u002F429, paywall, WAF, bot wall.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fweb\u002Fblocked-page-recovery\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":18},"01M7380VZ8P60Q64J3F92D4AMY","test-driven-development","TDD: enforce RED-GREEN-REFACTOR, tests before code.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fsoftware-development\u002Ftest-driven-development\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M7380QXN6F90R0MH1M3FHRKX","systematic-debugging","4-phase root cause debugging: understand bugs before fixing.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fsoftware-development\u002Fsystematic-debugging\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":18},"01M7380N4BZYT2WRRZN1N9DZ65","spike","Throwaway experiments to validate an idea before build.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fsoftware-development\u002Fspike\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":18},"01M7380JAEC9ZH5DVPHHK56P97","simplify-code","Parallel 4-agent cleanup of recent code changes.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fsoftware-development\u002Fsimplify-code\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":18},"01M7380F5MBB7FKWWV3G3TZ193","requesting-code-review","Pre-commit review: security scan, quality gates, auto-fix.","https:\u002F\u002Fgithub.com\u002Fashneil12\u002Fvanilla-hermes-agent-canary\u002Fblob\u002Fmain\u002Fskills\u002Fsoftware-development\u002Frequesting-code-review\u002FSKILL.md",[],[]]