chore: scaffold M0 — Rust engine crate, CLI, plugin, toolchain, CI#1
Open
amondnet wants to merge 5 commits into
Open
chore: scaffold M0 — Rust engine crate, CLI, plugin, toolchain, CI#1amondnet wants to merge 5 commits into
amondnet wants to merge 5 commits into
Conversation
Set up the adaptive-fetch project skeleton per docs/rfcs/0001-adaptive-fetch.md. - engine-src/: Rust workspace + crate with the stable public contract (fetch(), FetchResult, Verdict, FetchOptions) and stubbed engine-stage modules (transport/validators/scheduler/waf_detector/phase0/executor/ learning/safety/url_transforms) to fill across milestones M1–M5 - CLI `adaptive-fetch <URL> [--selector --device --trace --json]` with the R6 "NOT EXHAUSTED" failure-gate output; exit 0=success / 1=failure - mise.toml (rust+bun toolchain, fmt/clippy/test tasks), orca.yaml (rust worktree setup), .worktreeinclude, .gitignore, rustfmt.toml - .claude-plugin/plugin.json + skills/adaptive-fetch/SKILL.md stub - CI (.github/workflows/ci.yml): fmt + clippy -D warnings + build + test, third-party actions pinned to commit SHAs Engine returns an honest "unimplemented" result for now; network stages land in follow-up milestones.
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (9)
✅ Files skipped from review due to trivial changes (7)
📝 WalkthroughWalkthroughAdds the adaptive-fetch RFC, Rust engine scaffold, doc-only stage modules, and repository tooling, workflow, and policy files. The engine currently exposes public options/result types, a stub Changesadaptive-fetch scaffold and docs
Estimated code review effort: 4 (Complex) | ~60 minutes 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
🧹 Nitpick comments (1)
orca.yaml (1)
17-17: 🩺 Stability & Availability | 🔵 TrivialConsider documenting the
@pleaseai/worktreeincludepackage and its purpose.Line 17 uses
bunx@pleaseai/worktreeinclude`` but lacks inline documentation explaining what this package does. Adding a comment describing its role (e.g., "copies gitignored files to the worktree") would improve clarity for future maintainers.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@orca.yaml` at line 17, Add an inline or preceding comment to the line containing the `mise exec -- bunx `@pleaseai/worktreeinclude`` command that clearly describes the purpose and functionality of the `@pleaseai/worktreeinclude` package, such as explaining that it copies gitignored files to the worktree. This comment should provide enough context for future maintainers to understand why this package is being executed and what role it plays in the workflow.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In @.github/workflows/ci.yml:
- Line 20: The actions/checkout action on line 20 is missing the
persist-credentials parameter, which leaves the workflow token in git config and
creates a security vulnerability. Add persist-credentials: false as a parameter
to the actions/checkout action to ensure credentials are not persisted in the
git configuration for subsequent workflow steps, thereby hardening the CI
security posture.
In `@engine-src/src/lib.rs`:
- Around line 56-63: The FetchResult initialization lists "M4: Playwright
fallback" in the untried_routes vector, but sets must_invoke_playwright_mcp to
false. This inconsistency violates the failure-gate contract by failing to
signal that Playwright should be invoked. Change must_invoke_playwright_mcp from
false to true to correctly indicate that Playwright remains as an untried
fallback option and should be escalated by the CLI when needed.
In `@engine-src/src/result.rs`:
- Around line 80-82: The TODO comment above the `#[serde(skip)]` attribute on
the `content` field is stale and misleading. The comment claims that excluding
the field from JSON output is a TODO for M1, but the `#[serde(skip)]` macro
attribute already implements this behavior. Update or remove the TODO
designation from the comment to accurately reflect that the feature is already
implemented. You may keep the descriptive part about large HTML bloating trace
output if it provides useful context, but remove the TODO(M1) reference since it
no longer applies.
- Around line 43-49: The is_terminal_nonsuccess method incorrectly classifies
Verdict::RateLimited as terminal non-success, which contradicts the
documentation on line 27 that identifies RateLimited as transient. Remove
Verdict::RateLimited from the matches statement in the is_terminal_nonsuccess
method so that 429 rate-limited responses are treated as transient and allow
retries to continue, rather than prematurely stopping the retry loop.
---
Nitpick comments:
In `@orca.yaml`:
- Line 17: Add an inline or preceding comment to the line containing the `mise
exec -- bunx `@pleaseai/worktreeinclude`` command that clearly describes the
purpose and functionality of the `@pleaseai/worktreeinclude` package, such as
explaining that it copies gitignored files to the worktree. This comment should
provide enough context for future maintainers to understand why this package is
being executed and what role it plays in the workflow.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: a5827764-f014-4af4-a70c-2964071ef9bf
⛔ Files ignored due to path filters (1)
Cargo.lockis excluded by!**/*.lock
📒 Files selected for processing (25)
.claude-plugin/plugin.json.github/workflows/ci.yml.gitignore.worktreeincludeCargo.tomlREADME.mddocs/rfcs/0001-adaptive-fetch.mdengine-src/Cargo.tomlengine-src/src/executor.rsengine-src/src/learning.rsengine-src/src/lib.rsengine-src/src/main.rsengine-src/src/options.rsengine-src/src/phase0.rsengine-src/src/result.rsengine-src/src/safety.rsengine-src/src/scheduler.rsengine-src/src/transport.rsengine-src/src/url_transforms.rsengine-src/src/validators.rsengine-src/src/waf_detector.rsmise.tomlorca.yamlrustfmt.tomlskills/adaptive-fetch/SKILL.md
| name: fmt + clippy + build + test | ||
| runs-on: ubuntu-latest | ||
| steps: | ||
| - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2 |
There was a problem hiding this comment.
🔒 Security & Privacy | 🟠 Major | ⚡ Quick win
Disable credential persistence in checkout step.
Line 20 uses actions/checkout without persist-credentials: false; this leaves the workflow token in git config for later steps and weakens CI hardening.
Suggested patch
- - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+ - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+ with:
+ persist-credentials: false📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2 | |
| - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2 | |
| with: | |
| persist-credentials: false |
🧰 Tools
🪛 zizmor (1.26.1)
[warning] 20-21: credential persistence through GitHub Actions artifacts (artipacked): does not set persist-credentials: false
(artipacked)
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In @.github/workflows/ci.yml at line 20, The actions/checkout action on line 20
is missing the persist-credentials parameter, which leaves the workflow token in
git config and creates a security vulnerability. Add persist-credentials: false
as a parameter to the actions/checkout action to ensure credentials are not
persisted in the git configuration for subsequent workflow steps, thereby
hardening the CI security posture.
Source: Linters/SAST tools
Comment on lines
+56
to
+63
| untried_routes: vec![ | ||
| "M1: probe + 4-layer validation + SSRF transport".to_string(), | ||
| "M2: diversity grid scheduler + WAF detection".to_string(), | ||
| "M3: Phase 0 official-API router".to_string(), | ||
| "M4: Playwright fallback".to_string(), | ||
| ], | ||
| must_invoke_playwright_mcp: false, | ||
| } |
There was a problem hiding this comment.
🗄️ Data Integrity & Integration | 🟠 Major | ⚡ Quick win
Set must_invoke_playwright_mcp when Playwright remains untried.
fetch() advertises "M4: Playwright fallback" as untried, but Line 62 sets must_invoke_playwright_mcp to false. That breaks the FetchResult failure-gate contract and suppresses the CLI’s explicit MCP escalation hint.
Proposed fix
untried_routes: vec![
"M1: probe + 4-layer validation + SSRF transport".to_string(),
"M2: diversity grid scheduler + WAF detection".to_string(),
"M3: Phase 0 official-API router".to_string(),
"M4: Playwright fallback".to_string(),
],
- must_invoke_playwright_mcp: false,
+ must_invoke_playwright_mcp: true,📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| untried_routes: vec![ | |
| "M1: probe + 4-layer validation + SSRF transport".to_string(), | |
| "M2: diversity grid scheduler + WAF detection".to_string(), | |
| "M3: Phase 0 official-API router".to_string(), | |
| "M4: Playwright fallback".to_string(), | |
| ], | |
| must_invoke_playwright_mcp: false, | |
| } | |
| untried_routes: vec![ | |
| "M1: probe + 4-layer validation + SSRF transport".to_string(), | |
| "M2: diversity grid scheduler + WAF detection".to_string(), | |
| "M3: Phase 0 official-API router".to_string(), | |
| "M4: Playwright fallback".to_string(), | |
| ], | |
| must_invoke_playwright_mcp: true, | |
| } |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@engine-src/src/lib.rs` around lines 56 - 63, The FetchResult initialization
lists "M4: Playwright fallback" in the untried_routes vector, but sets
must_invoke_playwright_mcp to false. This inconsistency violates the
failure-gate contract by failing to signal that Playwright should be invoked.
Change must_invoke_playwright_mcp from false to true to correctly indicate that
Playwright remains as an untried fallback option and should be escalated by the
CLI when needed.
Comment on lines
+43
to
+49
| /// "Stop the grid — more TLS attempts cannot help." 429 is included here | ||
| /// (don't hammer) but is surfaced as a *transient* route by the failure gate. | ||
| pub fn is_terminal_nonsuccess(self) -> bool { | ||
| matches!( | ||
| self, | ||
| Verdict::AuthRequired | Verdict::NotFound | Verdict::RateLimited | ||
| ) |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟠 Major | ⚡ Quick win
Align RateLimited terminality with the transient 429 contract.
Line 27 documents RateLimited as transient, but Lines 45-49 classify it as terminal non-success. That will prematurely stop retries once this helper drives scheduler control flow.
Suggested fix
pub fn is_terminal_nonsuccess(self) -> bool {
matches!(
self,
- Verdict::AuthRequired | Verdict::NotFound | Verdict::RateLimited
+ Verdict::AuthRequired | Verdict::NotFound
)
}🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@engine-src/src/result.rs` around lines 43 - 49, The is_terminal_nonsuccess
method incorrectly classifies Verdict::RateLimited as terminal non-success,
which contradicts the documentation on line 27 that identifies RateLimited as
transient. Remove Verdict::RateLimited from the matches statement in the
is_terminal_nonsuccess method so that 429 rate-limited responses are treated as
transient and allow retries to continue, rather than prematurely stopping the
retry loop.
Comment on lines
+80
to
+82
| /// Fetched body. TODO(M1): exclude from `--json` (emit `content_length` | ||
| /// instead) so large HTML never bloats the trace output. | ||
| #[serde(skip)] |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Update the stale TODO on content JSON behavior.
Lines 80-82 say JSON exclusion is TODO(M1), but #[serde(skip)] already implements it. This comment is now misleading.
Suggested fix
- /// Fetched body. TODO(M1): exclude from `--json` (emit `content_length`
- /// instead) so large HTML never bloats the trace output.
+ /// Fetched body. Excluded from `--json`; use `content_length()` for
+ /// lightweight reporting in structured output.📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| /// Fetched body. TODO(M1): exclude from `--json` (emit `content_length` | |
| /// instead) so large HTML never bloats the trace output. | |
| #[serde(skip)] | |
| /// Fetched body. Excluded from `--json`; use `content_length()` for | |
| /// lightweight reporting in structured output. | |
| #[serde(skip)] |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@engine-src/src/result.rs` around lines 80 - 82, The TODO comment above the
`#[serde(skip)]` attribute on the `content` field is stale and misleading. The
comment claims that excluding the field from JSON output is a TODO for M1, but
the `#[serde(skip)]` macro attribute already implements this behavior. Update or
remove the TODO designation from the comment to accurately reflect that the
feature is already implemented. You may keep the descriptive part about large
HTML bloating trace output if it provides useful context, but remove the
TODO(M1) reference since it no longer applies.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
1 issue found across 26 files
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name=".github/workflows/ci.yml">
<violation number="1" location=".github/workflows/ci.yml:20">
P2: Set `persist-credentials: false` on the checkout step. Without it, the workflow token remains in `.git/config` and is accessible to all subsequent steps — a security hardening gap, especially since this workflow runs third-party actions (mise-action).</violation>
</file>
Architecture diagram
sequenceDiagram
participant Dev as Developer
participant CLI as adaptive-fetch binary
participant Engine as engine lib (adaptive_fetch)
participant CI as GitHub Actions CI
participant Tool as mise / cargo
Note over Dev,Engine: M0 Stub Flow
Dev->>CLI: adaptive-fetch <URL> [--json] [--trace]
CLI->>CLI: Parse args via clap
CLI->>Engine: fetch(url, &FetchOptions)
Engine-->>CLI: FetchResult { ok=false, stop_reason="unimplemented", untried_routes=[...] }
alt --json flag
CLI->>CLI: Serialize FetchResult to JSON (content excluded)
CLI-->>Dev: Print JSON to stdout
else default text output
CLI->>CLI: Print verdict, stop reason, content length
opt --trace
CLI->>CLI: Print each Attempt from trace
end
Note over CLI: R6 failure gate
CLI->>CLI: Print "NOT EXHAUSTED" block with untried_routes
end
CLI-->>Dev: Exit code 1
Note over Dev,CI: CI Pipeline (on push/PR)
CI->>Tool: mise install (rust + bun)
Tool-->>CI: Toolchain ready
CI->>Tool: cargo fmt --all -- --check
Tool-->>CI: format ok
CI->>Tool: cargo clippy --all-targets -- -D warnings
Tool-->>CI: clippy ok
CI->>Tool: cargo build --locked
Tool-->>CI: binary built
CI->>Tool: cargo test --locked
Tool-->>CI: 3 tests pass
CI-->>CI: Job completes
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
| name: fmt + clippy + build + test | ||
| runs-on: ubuntu-latest | ||
| steps: | ||
| - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2 |
There was a problem hiding this comment.
P2: Set persist-credentials: false on the checkout step. Without it, the workflow token remains in .git/config and is accessible to all subsequent steps — a security hardening gap, especially since this workflow runs third-party actions (mise-action).
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At .github/workflows/ci.yml, line 20:
<comment>Set `persist-credentials: false` on the checkout step. Without it, the workflow token remains in `.git/config` and is accessible to all subsequent steps — a security hardening gap, especially since this workflow runs third-party actions (mise-action).</comment>
<file context>
@@ -0,0 +1,26 @@
+ name: fmt + clippy + build + test
+ runs-on: ubuntu-latest
+ steps:
+ - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+ # Installs the toolchain pinned in mise.toml ([tools] rust + bun).
+ - uses: jdx/mise-action@e6a8b3978addb5a52f2b4cd9d91eafa7f0ab959d # v4.2.0
</file context>
chore: normalize LICENSE copyright to Passion Factory and add community-health files
Greptile SummaryThis PR scaffolds M0 of
Confidence Score: 5/5Safe to merge — this is a pure scaffold with no live network code; the only runnable logic is unit tests against stub return values. All engine stages are explicitly stubbed and return a fixed unimplemented result, so there is no behavioural regression risk. The public contract types are straightforward, the CLI is wired correctly, and CI enforces clippy with -D warnings and --locked builds. Both findings are design-level observations that do not affect the correctness of the scaffold. engine-src/src/options.rs has a minor dependency-graph concern worth addressing before M1 adds real library consumers. Important Files Changed
Flowchart%%{init: {'theme': 'neutral'}}%%
flowchart TD
CLI["CLI: adaptive-fetch URL\n--selector --device --trace --json"]
PARSE["Clap: parse args to FetchOptions"]
FETCH["fetch(url, opts) in lib.rs"]
STUB["M0 stub: ok=false, stop_reason=unimplemented"]
JSON_OUT["--json: FetchResult to stdout"]
HUMAN_OUT["non-json: content to stdout, summary+gate to stderr"]
TRACE["--trace: per-attempt diagnostics to stderr"]
EXIT["exit 0 (ok) / exit 1 (failure)"]
CLI --> PARSE --> FETCH --> STUB
STUB --> JSON_OUT --> EXIT
STUB --> HUMAN_OUT --> EXIT
STUB --> TRACE --> EXIT
%%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%%
flowchart TD
CLI["CLI: adaptive-fetch URL\n--selector --device --trace --json"]
PARSE["Clap: parse args to FetchOptions"]
FETCH["fetch(url, opts) in lib.rs"]
STUB["M0 stub: ok=false, stop_reason=unimplemented"]
JSON_OUT["--json: FetchResult to stdout"]
HUMAN_OUT["non-json: content to stdout, summary+gate to stderr"]
TRACE["--trace: per-attempt diagnostics to stderr"]
EXIT["exit 0 (ok) / exit 1 (failure)"]
CLI --> PARSE --> FETCH --> STUB
STUB --> JSON_OUT --> EXIT
STUB --> HUMAN_OUT --> EXIT
STUB --> TRACE --> EXIT
Reviews (2): Last reviewed commit: "chore: gitignore .claude/settings.local...." | Re-trigger Greptile |
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
| # optional npm-distributed wrapper. The engine itself needs no JS. | ||
| [tools] | ||
| rust = { version = "1.94", components = "rustfmt,clippy", profile = "minimal" } | ||
| bun = "latest" |
There was a problem hiding this comment.
bun = "latest" is not pinned — any new bun release will silently change the toolchain for every developer and in CI. Rust is correctly pinned to 1.94; bun should follow the same pattern with a specific version number.
Suggested change
| bun = "latest" | |
| bun = "1.2" |
Prompt To Fix With AI
This is a comment left during a code review.
Path: mise.toml
Line: 8
Comment:
The file is titled "Pinned local toolchain" but `bun = "latest"` is not pinned — any new bun release will silently change the toolchain for every developer and in CI. Rust is correctly pinned to `1.94`; bun should follow the same pattern with a specific version number.
```suggestion
bun = "1.2"
```
How can I resolve this? If you propose a fix, please make it concise.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
Comment on lines
+80
to
+83
| /// Fetched body. TODO(M1): exclude from `--json` (emit `content_length` | ||
| /// instead) so large HTML never bloats the trace output. | ||
| #[serde(skip)] | ||
| pub content: String, |
There was a problem hiding this comment.
--json (emit content_length instead)", but #[serde(skip)] already omits content from JSON. What isn't done yet is adding a serialized content_length field to the JSON output. The comment should reflect what still needs to land in M1 to avoid making future implementers think the work is already complete.
Suggested change
| /// Fetched body. TODO(M1): exclude from `--json` (emit `content_length` | |
| /// instead) so large HTML never bloats the trace output. | |
| #[serde(skip)] | |
| pub content: String, | |
| /// Fetched body. Excluded from `--json` via `#[serde(skip)]`. | |
| /// TODO(M1): add a serialized `content_length` field so JSON consumers | |
| /// can inspect body size without the full text bloating the trace output. | |
| #[serde(skip)] | |
| pub content: String, |
Prompt To Fix With AI
This is a comment left during a code review.
Path: engine-src/src/result.rs
Line: 80-83
Comment:
The TODO says "exclude from `--json` (emit `content_length` instead)", but `#[serde(skip)]` already omits `content` from JSON. What isn't done yet is adding a serialized `content_length` field to the JSON output. The comment should reflect what still needs to land in M1 to avoid making future implementers think the work is already complete.
```suggestion
/// Fetched body. Excluded from `--json` via `#[serde(skip)]`.
/// TODO(M1): add a serialized `content_length` field so JSON consumers
/// can inspect body size without the full text bloating the trace output.
#[serde(skip)]
pub content: String,
```
How can I resolve this? If you propose a fix, please make it concise.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
Keep internal marketplace config and telemetry repo labels out of upstream — the file is personal (deep-merged by pleaseai/oss apply-settings.sh) and must never reach the public history.
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
M0 project scaffold for
adaptive-fetch— a resilient, site-agnostic public-page reader for Claude Code, with the engine in Rust (native browser TLS-fingerprint impersonation). Engine stages are stubbed; this milestone establishes the crate, CLI, plugin packaging, toolchain, and CI. Design:docs/rfcs/0001-adaptive-fetch.md.What's included
fetch(),FetchResult,Verdict,FetchOptions) and stubbed engine-stage modules (transport / validators / scheduler / waf_detector / phase0 / executor / learning / safety / url_transforms) to fill across milestones M1–M5.adaptive-fetch <URL> [--selector --device --trace --json]with the R6 "NOT EXHAUSTED" failure-gate output; exit0=validated success /1=failure.mise.toml(rust + bun, fmt/clippy/test/check tasks),orca.yaml(rust worktree setup),.worktreeinclude,.gitignore,rustfmt.toml..claude-plugin/plugin.json+skills/adaptive-fetch/SKILL.mdstub..github/workflows/ci.yml: fmt + clippy (-D warnings) + build + test, third-party actions pinned to commit SHAs.Status
Engine currently returns an honest
stop_reason="unimplemented"result; network stages land in follow-up milestones (M1: probe + 4-layer validation + SSRF transport).Verification
Locally:
cargo fmt --check,cargo clippy -D warnings,cargo build --locked,cargo testall pass (3 tests). CLI smoke test prints the failure-gate block and exits 1.Notes
maindirectly to respect review workflow.Summary by cubic
Scaffolded M0 for adaptive-fetch: a Rust engine crate, CLI, Claude Code plugin, dev toolchain, CI, and community health files. Establishes the stable
fetch()contract; current runs returnstop_reason="unimplemented"(no network yet).FetchResult,Verdict,FetchOptions, and stubbed stages (transport/validators/scheduler/waf_detector/phase0/executor/learning/safety/url_transforms).adaptive-fetch <URL> [--selector --device --trace --json]with exit0on validated success,1on failure, and an R6 failure gate that prints "NOT EXHAUSTED"..claude-plugin/plugin.jsonandskills/adaptive-fetch/SKILL.md; design indocs/rfcs/0001-adaptive-fetch.md.mise.toml,orca.yaml, and a CI workflow running fmt, clippy (-D warnings), build, and test with actions pinned to SHAs..gitignoreupdated to exclude.claude/settings.local.json.Written for commit b1f6443. Summary will update on new commits.
Summary by CodeRabbit
adaptive-fetchClaude Code plugin and initial Rust engine/CLI scaffold for adaptive public-page fetching with traceable results and clear “not implemented” exit behavior.