Octos is a self-hosted AI assistant that runs on a user's computer and can connect to different AI providers, execute code, browse the web, remember information, schedule jobs, and create documents through browser, terminal, and messaging interfaces. The catalogue entries extend or configure this assistant's workflows.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add octos-org/octos --skill pptxgit clone --depth 1 https://github.com/octos-org/octosWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/octos-org/octos/pptx)<a href="https://agentmods.dev/skills/octos-org/octos/pptx"><img src="https://agentmods.dev/badge/skills/octos-org/octos/pptx/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/octos-org/octos/pptx"><img src="https://agentmods.dev/badge/skills/octos-org/octos/pptx.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00075 | $0.02673 |
| Opus 5.5 | $0.00030 | $0.01069 |
| Sonnet 5.5 | $0.00015 | $0.00535 |
| Haiku 4.5 | $0.00007 | $0.00267 |
Grade A, and why
pptx scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
可编辑 PPTX
制作模型负责内容、设计、看图与修复;独立评审会话负责复核实际交付稿。工具负责文件、字体计算、实际渲染、测量和证据整理;工具成功或检查项全绿不等于视觉通过。普通文案、解释或只读检查不自动启动整稿重做。
新环境首次制作时先按 portable runtime 检查依赖,
用随包的 scripts/setup_runtime.py 准备工作区环境;技能目录来自当前 catalog。
不依赖已有评测目录,不把安装完成称为视觉验收通过。
制作与返修必须使用 peer,并落实退出标准
制作、重做或修复 PPTX 成品时,采用“独立评审主会话+制作 peer”;用户明确指定其他模式时以用户要求为准。启动前读 peer 评审与退出。Octos 必须实际调用 peer_handoff,保存真实会话回执并确认双方能直接看图;不能以同一模型自评、普通后台脚本或仅写了分工说明冒充 peer。
制作 peer 独占改稿;评审者查看实际 PPTX 渲染,按证据把缺陷交回同一个 peer。先约定必修项、可选微调、停滞窗口和用户已有预算。必修项关闭且整稿复核通过就交付;连续无实质改善就退出当前修法;真实阻塞或预算耗尽时交回未完成结果。偏好性微调不能无限重开返修。
Octos 制作 peer 同时按 continuation evidence 初始化并核验现有退出策略。技能已安装、policy 文件存在、peer 已启动,都不证明内核已消费检查证据。 首个完整修复检查周期后必须确认运行时状态与该证据一致;缺口要明确报告并处理,不能继续声称阈值已启用。保留已运行任务的历史,不静默改写冻结实验。
先确定设计路径
- 从文稿新建或重做图文丰富的稿件:先读 视觉设计与验收,把论点转成整页的图形、关系与标签布局,再进入构建。用户指定 image-first/海报式全页生图时,按 成对设计 生成完整有字设计、从同一设计派生无字底稿,再重建可编辑文字。生成几个局部插画再塞入固定栏位,不等于完成整页设计。
- 已有被认可的设计图与无字底稿:沿用其构图,进入下面的原生叠字流程;不用重复设计或生图。
- 用户明确要求图形、表格也全部可编辑:选择适合的原生图形/图表实现,同时遵守整页视觉要求。单说“可编辑 PPTX”不自动要求每个像素都矢量化,也不授权改变指定的生图路径;说明实际可编辑范围。
用户要求改变信息密度、图文关系或版式时,重新检查整页构图与生图提示词;保持旧版位、仅换插图或缩字,不能证明已满足这个修改。两栏本身不是错误,是否适用由本页论点与用户要求决定。
保持本任务选定的流程
开始制作时,在项目已有的约定/制作记录中简短记下:内容来源与范围、选定路径、参考设计、可编辑范围、字体与分辨率要求。记录当前产物、尚未完成的步骤和真实阻塞;不另建一套调度或验收框架。开始、恢复任务或接到局部修改时核对这份记录。用户说“继续”、改字体或修一页时,原有范围与流程继续有效;用户明确改变要求时更新相关约定。
对已选定的 image-first 路径,有字整页图是布局参考;从它去字后叠原生文字,才能同时保留设计与可编辑性。不能以“图片里的字不能编辑”为由改成只生成无字插画、固定栏位排版或只交整页图片。已有合适的有字设计可直接复用。只需复用仍然有效的步骤;只有实际输入缺失、工具失败或不支持的能力才构成阻塞,排版困难本身不授权更换流程。
模型决定构图、提示词、标签位置和修复方式。实现可以调整,但须保持用户选定路径及其可核对的产物关系。分批完成时更新进度,继续处理已获授权的剩余工作;进度汇报不等于任务完成,不逐批重新请求“继续”的许可。
常用制作路径
- 确定内容版本、页数、比例、字体和可编辑范围。最新文稿决定说什么,用户认可的参考决定如何画。实际查看参考图;OCR 辅助定位,不能替代原文。保留被指定的旧版。
- 对“无字插画位图+原生文字”,使用 native builder,只编写页面数据。共享字体与构建实现复用一次。独立视觉对象的标签分别定位;不要把多个卡片的标签塞进一个靠空行对齐的文本框。 遇到倾斜铭牌、斜向文字或透视面,按 标签几何 测量具体标签四角,重建局部文字框及旋转;检查字形与斜边、基线的关系。整个场景面板不能替代局部标签边界,旋转也不等于透视变形。
- 字体、换行或可见字形定位需要计算时,用 字体提案与修订 中的
native_fit.py。明确实际字体文件;--ink-align可按字形边界和基线提议文本框位置。提案不改原稿,未明确允许缩小时不自动缩字;布局归属由模型根据图形判断。 native_cycle.py run --inspect合并构建、全稿/无字稿渲染和测量,返回对象级问题、局部图和归属叠图。按 共同页面方法 批量修复;native_cycle.py revise返回前后对照及仍存在/新增/不再检出的问题,--resume继续同一个异步任务。未改动对象的问题仍需判断;边界变化不等于文字修好了。- 按 三项独立核对 检查最终 PPTX:原稿与原生文字的完整意思、有字设计与无字底图的图形变化、实际渲染的阅读与标签归属。另按 视觉验收 比较整页图文关系、信息密度和全稿节奏。全页检查不局限于检测器标记的对象;密集区域用清晰的上下文裁图。检查结果中的未知项保留为未知,复用未改变部分的有效证据。制作 peer 自查后交回独立评审会话,按退出标准完成,不要求用户逐轮批准。
What ships with it
60 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 322 B
- references/checked-jobs.md 2.5 KB
- references/continuation-evidence.md 6.4 KB
- references/efficient-production.md 6.1 KB
- references/fit-and-revise.md 5.5 KB
- references/host-checks.md 6.1 KB
- references/independent-pages.md 23 KB
- references/label-geometry.md 3.8 KB
- references/native-builder.md 13 KB
- references/ocr-artifact-sweep.md 8.4 KB
- references/page-method.md 6.1 KB
- references/page-worker.md 2.8 KB
- references/paired-workflow.md 9.7 KB
- references/peer-review-exit.md 6.1 KB
- references/pipeline-pitfalls.md 5.8 KB
- references/portable-runtime.md 3.1 KB
- references/powerpoint-compatibility.md 2.7 KB
- references/pptx-contract.md 8.4 KB
- references/pptx-first-audit.md 5.2 KB
- references/reconstruction-repair.md 6.1 KB
- references/review-evidence.md 6.2 KB
- references/text-layout-reinspection.md 11 KB
- references/text-region-audit.md 7.2 KB
- references/visual-review.md 7.0 KB
- requirements.txt 110 B
- scripts/audit_text_regions.py 13 KB runs code
- scripts/check_pptx_compat.py 3.1 KB runs code
- scripts/checked_jobs.py 12 KB runs code
- scripts/compare_slide_proof.py 4.6 KB runs code
- scripts/continuation_evidence.py 12 KB runs code
- scripts/copy_glossary.py 1.2 KB runs code
- scripts/defect_followup.py 6.5 KB runs code
- scripts/inspect_rasters.py 2.0 KB runs code
- scripts/inspect_slide_layers.py 6.7 KB runs code
- scripts/label_geometry.py 6.8 KB runs code
- scripts/local_tools.py 12 KB runs code
- scripts/native_cycle.py 15 KB runs code
- scripts/native_deck.py 27 KB runs code
- scripts/native_fit.py 18 KB runs code
- scripts/native_inspect.py 47 KB runs code
- scripts/native_review.py 3.1 KB runs code
- scripts/native_revise.py 7.2 KB runs code
- scripts/operation_receipts.py 3.4 KB runs code
- scripts/paired_deck.py 11 KB runs code
- scripts/pptx_check_client.py 1.7 KB runs code
- scripts/pptx_contract.py 16 KB runs code
- scripts/pptx_host_check.py 28 KB runs code
- scripts/prepare_page_jobs.py 6.9 KB runs code
- scripts/raster_overlap.py 5.2 KB runs code
- scripts/render_cache.py 12 KB runs code
- scripts/render_deck.py 4.0 KB runs code
- scripts/review_cache.py 772 B runs code
- scripts/review_summary.py 1.1 KB runs code
- scripts/review_transport.py 6.1 KB runs code
- scripts/setup_runtime.py 4.0 KB runs code
- scripts/test_copy_glossary.py 966 B runs code
- scripts/test_defect_followup.py 2.6 KB runs code
- scripts/tests/test_checked_jobs.py 14 KB runs code
- scripts/tests/test_continuation_evidence.py 8.3 KB runs code
- scripts/tests/test_deck_assembly.py 9.7 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago Changed · +9 lines ec2c9099c5b0
- 6d ago First seen · 67 lines · 75 tokens per session scan A 2d9119318dd1
pptx is a skill published in the GitHub repository octos-org/octos (766 stars, last pushed today), licensed Apache-2.0. It adds 75 tokens to every session and 2,673 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-31.
pptx compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| pptxoctos-org/octos | Skill | 766 | today | Apache-2.0 | A | 75 | Codex |
| orbit-notionnexu-io/open-design | Skill | 106k | today | Apache-2.0 | A | 117 | Claude Code, Codex |
| instrument-data-to-allotropeanthropics/knowledge-work-plugins | Skill | 35k | 2d ago | Apache-2.0 | A | 123 | Claude Code |
| apify-youtube-transcript-scrapersickn33/agentic-awesome-skills | Skill | 49k | yesterday | MIT | A | 48 | Claude Code |
| legal-holdanthropics/claude-for-legal | Skill | 9.9k | 3d ago | Apache-2.0 | A | 68 | Claude Code |
| ppt-reviewOrkas-AI/Orkas | Skill | 9.0k | today | MIT | A | 2 | Claude Code, Codex |
| feishuHmbown/CodeWhale | Skill | 41k | 5d ago | MIT | A | 33 | Claude Code, Codex |