GitHub’s page for Binaryify/open-kimi-ppt-skill says the repository was archived and made read-only after its contents were cleared for copyright reasons. That removal is the central fact about the project now: its listing still promises an unofficial Kimi Slides skill for AI agents, editable PPTD and PPTX output, and a local browser editor, but the repository provides no code with which to inspect those promises.
The project’s visibility has outlasted its implementation. GitHub displays 1.6k stars and 1.2k forks on the page, while the repository file listing shows only README.md and a single commit. The README states, in Chinese and English, that all repository content has been cleared because of copyright reasons.
That makes this less a story about a usable presentation-generation tool than about the limits of repository metadata. A project description can remain searchable and attractive long after the material needed to evaluate its architecture, licensing, installation path, and output quality has disappeared.
The advertised workflow cannot be inspected
The repository’s About text describes an “Unofficial Kimi Slides skill for AI agents” that can generate editable PPTD and PPTX files and includes a local browser editor. Those are meaningful claims, not cosmetic features. “Editable” could distinguish a structured presentation artifact from a flattened image export; a browser editor could imply an iteration path after generation; support for two formats could affect interoperability.
None of those implications can be confirmed from the remaining repository material.
There is no visible implementation to establish what a “skill” meant in this project, how an agent invoked it, whether PPTD and PPTX were generated directly or through conversion, or what editing operations the local editor supported. There is also no surviving documentation in the visible README beyond the clearance notice. The project page therefore supports reporting the claimed scope, but not reporting the scope as demonstrated capability.
That distinction matters especially for developer readers assessing tools that sit between generative systems and familiar office formats. A slide deck that opens in a presentation application is not necessarily a deck whose text, layouts, charts, or assets remain practical to edit. Likewise, a local editor may be a substantial authoring environment or merely a narrow preview surface. With the source removed, GitHub’s short description cannot resolve those questions.
The sober reading is that the page preserves an assertion, not evidence for the assertion.
Copyright clearance changes the engineering assessment
The stated reason for removal is copyright. The page does not explain what material raised the issue, who asserted rights, whether the disputed material was source code, assets, documentation, or another dependency, or whether a replacement is planned. It would be speculation to assign a narrower cause.
Still, the outcome is concrete. The repository is archived, read-only, and cleared. A prospective user cannot obtain a normal source distribution from the visible project page, and a prospective contributor has no codebase there to review or improve. The six listed issues do not alter that position; the repository has no visible pull requests and its content has been removed.
For technical adoption, copyright removal is not merely a maintenance inconvenience. It breaks the normal chain of evidence that lets teams assess an open-source dependency: inspect the code, identify bundled assets and dependencies, review the license terms, reproduce a build, and determine whether the project can be maintained internally. The source page supplies none of that material now.
The remaining fork count should not be treated as a substitute. GitHub shows 1.2k forks, which suggests that copies may have existed across the network, but the page does not establish what any fork contains, whether it remains available, whether it is current, or whether it resolves the copyright problem. A fork count is evidence of historical interest, not a clean bill of technical or legal health.
Popularity metadata is now a misleading signal
The project’s 1.6k stars and 1.2k forks give the listing an appearance of maturity. In ordinary circumstances, those figures can help developers find active projects or gauge community attention. Here they are historical signals attached to a repository whose substantive contents no longer exist.
The page also has a concise product proposition that is easy to rediscover: generate presentation files through an AI-agent-oriented skill and edit them locally in a browser. Search results, repository cards, and saved links can continue to surface that proposition without making the clearance status equally prominent. The archived label and README notice correct the record for visitors who open the repository, but they do not restore the evidence behind the original claims.
This is a useful warning against reading GitHub engagement as proof of current availability. Stars do not prove a release can be built. Forks do not prove a package can be lawfully reused. An About field does not prove a format exporter preserves editability. Those are separate questions, and this project page now leaves all three unanswered.
The project should therefore be classified conservatively: it is an archived, cleared repository whose public metadata describes an AI-agent presentation workflow, rather than a verifiable implementation of one. Teams considering it as a reference, dependency, or integration target should not promote its advertised PPTD, PPTX, or browser-editing features into requirements until they can examine a legitimate, complete source distribution and its terms.
The unresolved trade-off is sharp: developers may value the preserved idea of editable, agent-generated slides, but the only visible official repository record no longer provides the code or rights context needed to turn that idea into a dependable engineering choice.
