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 dyngai/handoffkit --skill handoffkit-scaffoldgit clone --depth 1 https://github.com/dyngai/handoffkitWrote 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/dyngai/handoffkit/handoffkit-scaffold)<a href="https://agentmods.dev/skills/dyngai/handoffkit/handoffkit-scaffold"><img src="https://agentmods.dev/badge/skills/dyngai/handoffkit/handoffkit-scaffold/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/dyngai/handoffkit/handoffkit-scaffold"><img src="https://agentmods.dev/badge/skills/dyngai/handoffkit/handoffkit-scaffold.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 | $0.00108 | $0.01255 |
| Opus 5.5 | $0.00043 | $0.00502 |
| Sonnet 5.5 | $0.00022 | $0.00251 |
| Haiku 4.5 | $0.00011 | $0.00126 |
Grade C, and why
handoffkit-scaffold scanned grade C with 1 finding 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 2mo 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
is there. Do not overwrite, delete, merge into, or `rm -rf` those How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scaffold PlatformPilot-style HandoffKit primitives into Go
Use this to add the DynG AI (https://dyng.ai) HandoffKit message-passing
runtime to a Go project: the practical, PlatformPilot
(https://platformpilot.ai) inspired foundation for explicit agent handoffs,
inspectable routing, bounded fan-in, and traceable coordination. The reference
sources are vendored under this skill's references/_src/ directory, a
snapshot of the canonical runtime, so this works offline. (The _src name
keeps the go tool from compiling these reference files as part of the
HandoffKit module.)
This is the Go reference implementation, one instantiation of the pattern.
The pattern itself is language-agnostic: for a Python, TypeScript, Rust, or
Erlang project there is nothing to copy, use the handoffkit skill and build the
equivalent from its primitive→language mapping table.
Steps
-
Determine the target module path. Read the project's
go.mod. If there is none, ask the user for a module path and rungo mod init <path>. The scaffolded runtime requires Go 1.21+. The full HandoffKit repo requires Go 1.22+ because of its OpenAI SDK dependency. -
Choose package destinations and protect existing code. Before copying, inspect the target paths you plan to use.
- Default destinations are
sketch/andruntime/. - If either path already exists and contains files, stop and summarize what
is there. Do not overwrite, delete, merge into, or
rm -rfthose directories implicitly. - Ask the user whether to overwrite/merge into those paths, or choose an
alternate namespace such as
internal/handoffkit/sketchandinternal/handoffkit/runtime. - If you choose alternate paths, keep the same relative shape (
sketchbesideruntime) and rewrite imports to that alternate sketch package.
- Default destinations are
-
Copy the reference sources from this skill into the chosen destinations, preserving structure, tests, and license attribution:
references/_src/sketch/*.go→<sketch destination>/references/_src/runtime/*.go→<runtime destination>/- Include
_test.gofiles. They are part of the scaffolded confidence boundary and let the target project verify the copied runtime directly. - The copied files are MIT-licensed HandoffKit source. Preserve attribution
by either copying the bundled HandoffKit license from this repo's root
LICENSEfile alongside the scaffolded HandoffKit code, or creating/updating the target project's third-party notices file with the HandoffKit name, source URLhttps://github.com/dyngai/handoffkit, MIT license, and full license text from that bundled rootLICENSEfile.
What ships with it
27 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 306 B
- assets/symbol-square.png 20 KB
- references/_src/runtime/broker_test.go 9.5 KB
- references/_src/runtime/broker.go 2.9 KB
- references/_src/runtime/budget_test.go 3.6 KB
- references/_src/runtime/budget.go 2.8 KB
- references/_src/runtime/compactor_test.go 3.3 KB
- references/_src/runtime/compactor.go 4.8 KB
- references/_src/runtime/corpus_test.go 7.1 KB
- references/_src/runtime/corpus.go 4.6 KB
- references/_src/runtime/deadletter_test.go 6.2 KB
- references/_src/runtime/deadletter.go 3.6 KB
- references/_src/runtime/join_test.go 5.0 KB
- references/_src/runtime/join.go 2.5 KB
- references/_src/runtime/mailbox_test.go 3.5 KB
- references/_src/runtime/mailbox.go 3.1 KB
- references/_src/runtime/quorum_test.go 5.5 KB
- references/_src/runtime/quorum.go 3.6 KB
- references/_src/runtime/router_test.go 4.9 KB
- references/_src/runtime/router.go 5.3 KB
- references/_src/runtime/selector_test.go 6.7 KB
- references/_src/runtime/selector.go 4.3 KB
- references/_src/runtime/supervisor_test.go 14 KB
- references/_src/runtime/supervisor.go 8.2 KB
- references/_src/runtime/trace_test.go 5.4 KB
- references/_src/runtime/trace.go 1.6 KB
- references/_src/sketch/handoffkit.go 6.5 KB
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.
- 2mo ago First seen · 84 lines · 108 tokens per session scan C b55c8b1b47a5
handoffkit-scaffold is a skill published in the GitHub repository dyngai/handoffkit (31 stars, last pushed 4mo ago), licensed MIT. It adds 108 tokens to every session and 1,255 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
handoffkit-scaffold compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| handoffkit-scaffolddyngai/handoffkit | Skill | 31 | 4mo ago | MIT | C | 108 | Codex |
| neo4j-driver-go-skillneo4j-contrib/neo4j-skills | Skill | 121 | 3d ago | MIT | A | 143 | Claude Code |
| bensz-r-developerhuangwb8/skills | Skill | 50 | 3d ago | MIT | A | 125 | Claude Code, Codex |
| go-concurrency-reviewerRaNDoM6913/claude-code-superkit | Skill | 3 | yesterday | MIT | A | 35 | Claude Code |
| go-error-reviewerRaNDoM6913/claude-code-superkit | Skill | 3 | yesterday | MIT | A | 29 | Claude Code |
| go-modernizerRaNDoM6913/claude-code-superkit | Skill | 3 | yesterday | MIT | A | 32 | Claude Code |
| go-performance-reviewerRaNDoM6913/claude-code-superkit | Skill | 3 | yesterday | MIT | A | 34 | Claude Code |