chore(seo): broader crates.io framing — spend, risk, audit#31
Merged
Conversation
Cycles is the runtime authority for AI agents across three problems
that observation alone can't fix: spend, risky tool actions, and
audit gaps. Previous metadata and README opening focused only on
the spend dimension, which understated what the crate actually does
and missed half the search-intent keyword pool.
Cargo.toml:
description: "Rust client for the Cycles budget-management
protocol — deterministic spend control for AI agents
and LLM workflows"
→ "Runtime authority for AI agents in Rust — hard
limits on agent spend, risky tool actions, and audit
gaps. Tokio-native client for the Cycles protocol
(reserve-commit lifecycle, RAII guards)."
keywords: ["cycles", "budget", "llm", "ai-agents", "cost-control"]
→ ["ai-agents", "llm", "budget", "governance", "audit-log"]
Why the keyword swap:
- "cycles" is brand-only; people who know the brand will find
the crate by name regardless. Wasted slot.
- "cost-control" is a niche compound term with low search.
- "governance" captures the action-authority dimension and
compliance-adjacent searches.
- "audit-log" captures the audit-trail dimension explicitly.
- Kept "ai-agents", "llm", "budget" — the three highest-volume
relevant search terms.
README.md:
H1 broadened from "AI agent budget and action authority SDK" to
"Runtime Authority for AI Agents (Spend, Actions, Audit)".
Opening section rewritten to introduce all three pillars
(spend / risky actions / audit gaps) with a short bullet for each,
rather than leading with budget enforcement and treating the
others as secondary. Clarifies the Decision model (Allow /
AllowWithCaps / Deny) up front and notes that the same wire
protocol is shared with Python / TypeScript / Spring Boot clients.
AUDIT.md:
Date line gets a third entry noting today's metadata refresh.
No behavioral or conformance changes; protocol audit results
unchanged.
Verified: cargo build clean. No code changes — metadata and
documentation only.
Merged
2 tasks
amavashev
added a commit
that referenced
this pull request
May 8, 2026
Releases the broader crates.io framing landed in #31. No behavioral or API changes — metadata, README, and CHANGELOG only. Cargo.toml: version 0.2.3 → 0.2.4 CHANGELOG.md: new [0.2.4] - 2026-05-08 entry under "Changed" Cargo.lock: regenerated by cargo build with the new version After this PR merges, push the v0.2.4 tag to trigger the automated crates.io publish step in .github/workflows/ci.yml (on: push: tags: ["v*"] → "Publish to crates.io" job → cargo publish). Conformance audit results unchanged from 0.2.3 (95.55% coverage, all 9/9 endpoints + 6/6 request schemas + 10/10 response schemas match v0.1.24 spec).
amavashev
added a commit
to runcycles/cycles-docs
that referenced
this pull request
May 8, 2026
…udit A read-only audit of quickstart/ found 8 pages framing Cycles as a spend/budget tool, missing the other two pillars (risky tool actions and audit gaps). Earlier today we made the same correction in the cycles-client-rust crate metadata, README, and the related blog post (runcycles/cycles-client-rust#31, #32, #33; #585). This commit applies the same pattern to the public quickstart surface using a light-touch approach: - Frontmatter `description` rewritten on each page to name spend + action + audit (replaces "budget enforcement" framing) - One short ::: tip ::: callout inserted near the top of each page, listing the three pillars with concrete primitives — caps for risky actions, signed events for audit - No title changes, no restructuring, no code-example edits Pages updated (8): Light SDK pattern (4): same callout, identical except for SDK name - quickstart/getting-started-with-the-python-client.md - quickstart/getting-started-with-the-typescript-client.md - quickstart/getting-started-with-the-cycles-spring-boot-starter.md - quickstart/getting-started-with-the-rust-client.md MCP server (1): callout uses MCP tool names (cycles_decide, cycles_create_event) instead of generic primitives - quickstart/getting-started-with-the-mcp-server.md Explainer / strategy (3): same callout + small body tweaks - quickstart/what-is-cycles.md (one-line opener also broadened) - quickstart/how-to-add-hard-budget-limits-to-spring-ai-with-cycles.md - quickstart/how-to-choose-a-first-cycles-rollout-...md (callout adds a sentence noting which pillar each rollout option primarily addresses, since the page is about choosing scope) Pages intentionally NOT changed (audited and judged appropriate as-is): - quickstart/index.md (already cross-links to /protocol/, /how-to/) - quickstart/end-to-end-tutorial.md (runbook format) - quickstart/architecture-overview-* (technical reference) - quickstart/deploying-* and self-hosting-* (operations focus) - quickstart/mcp-claude-* + cursor + windsurf (setup guides; existing warning correctly clarifies MCP availability ≠ enforcement) Title and URL slug on each page intentionally preserved to keep inbound search-link equity. The two spend-only strategy pages (spring-ai how-to, rollout-choice) keep their existing slugs since those are the canonical search targets for "spring ai budget limit" and "cycles rollout" queries respectively. Diff shape: each file gets a description swap (1 line replaced) and a callout insertion (5-7 lines added). +59 insertions / -9 deletions total across 8 files. Verified: - 83/83 tests pass - npm run build succeeds (98s) - rendered HTML <meta name="description"> reflects new copy - rendered HTML contains the callout block exactly once per page
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Cycles is the runtime authority for AI agents across three problems that observation alone can't fix: spend, risky tool actions, and audit gaps. The previous Cargo.toml description and README opening leaned on the spend dimension only, which understated what the crate actually does and missed half the search-intent keyword pool on crates.io and Google.
This PR broadens the framing — no code changes, metadata + docs only.
Changes
Cargo.toml["cycles", "budget", "llm", "ai-agents", "cost-control"]["ai-agents", "llm", "budget", "governance", "audit-log"]Why the keyword swap:
cycles— brand-only. People who know the brand find the crate by name regardless. Wasted slot.cost-control— niche compound. Low search.governance— captures action-authority + compliance searches.audit-log— captures audit-trail searches explicitly.ai-agents,llm,budget— the three highest-volume relevant search terms.README.mdAUDIT.mdTest plan
cargo build— clean