docs(miner): inventory real-world contribution eligibility signals for the AMS profile

[luciferlive112116]

Summary

The ContributionProfile schema (#6795) must be grounded in what real repos actually expose, not an assumption. This audits 10 diverse public repos from the live GitHub API and writes the inventory to packages/loopover-miner/docs/ams-contribution-signal-inventory.md, following the ams-storage-abstraction-research.md / ams-auth-identity-research.md precedent in that same directory.

Research and docs only — no code, no schema, no discover wiring, exactly as the issue scopes.

The sample

Two JSONbored gate-enabled repos, six large OSS projects chosen for divergent contribution norms (react, rust, svelte, deno, tailwindcss, kubernetes), one repo with agent docs but no human contribution docs (sure-aio), and one with no contribution docs at all (sindresorhus/slugify — the negative case the issue explicitly asks for).

What the evidence says — and why it changes the schema

Every figure was read from GET /repos/{owner}/{repo}/labels and .../contents/{path} on 2026-07-17, not from memory. Four findings, each with a direct schema implication:

1. No universal eligibility label. good first issue / help wanted are present in only 6 of 10. The three largest by contributor activity — rust, deno, kubernetes — use their own taxonomies (rust: E-easy/E-medium/E-hard). A fixed-name matcher silently returns "no eligible issues" on exactly those repos.
2. Label descriptions are unreliable, unevenly. Coverage runs 3/17 (tailwind) and 8/76 (react) up to 99/100 (rust). The schema can't depend on the description — but can't ignore it either, since rust encodes the semantics only there (E-easy reads "…Good first issue.").
3. CONTRIBUTING.md is often a signpost. react's is 208 bytes, kubernetes' 525 — both just links to an external guide AMS can't read. "Has CONTRIBUTING.md" is a near-worthless presence check.
4. The linked-issue requirement is loopover-local, not an ecosystem norm. Grepping each real CONTRIBUTING.md: loopover 8 mentions, react/rust/kubernetes 0. Modeling it as a core profile field would encode our own gate's norm as if it were general — a mistake this doc exists to prevent the next issue from making.

Plus one genuinely surprising result: AGENTS.md/CLAUDE.md is the most consistent non-label signal (6/10, including react/deno/k8s/svelte), and sure-aio has agent docs but no CONTRIBUTING — so extraction that treats human docs as primary and agent docs as fallback has the priority backwards.

The doc closes with an explicit "what this means for #6795" section, so the schema-design issue can build against evidence rather than re-derive it.

Validation

• docs:drift-check — ok (the doc adds no config/command surface).
• prettier --check — clean (Markdown formatting).
• No forbidden public-surface terms — verified the doc contains no wallet/hotkey/coldkey/trust-score/reward wording, and deliberately does not quote loopover's own gittensor:* label reward multipliers even though they were in the raw data.
• Single-file diff under a wanted path (packages/loopover-miner/docs/); no code, so no coverage/CI-suite obligation; rebased on latest main, no base conflict.

Scope & safety

• One new documentation file. No source, no test, no schema, no discover change — the implementation sub-issues (AMS contribution-profile: design the ContributionProfile schema and caching shape #6795–AMS contribution-profile: wire eligibility filtering into discover #6798) build on this.
• Read-only research; every claim is reproducible from the public API, and the method/reproducibility section says exactly how.
• No secrets; no changelog, site/, CNAME, or lovable changes.

Closes #6794

[superagent-security]

Superagent didn't find any vulnerabilities or security issues in this PR.

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-17 17:12:21 UTC}

1 file · 1 AI reviewer · no blockers · readiness 98/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This is a docs-only research spike (no code, schema, or wiring changes) that audits 10 public repos' labels, CONTRIBUTING.md files, PR templates, and agent docs to ground the future ContributionProfile schema (#6795) in real-world data. The findings are specific and well-cited (exact label/description counts, byte sizes, grep counts), the negative case (slugify) and sample rationale are clearly stated, and the scope discipline (research only, explicitly no code/schema/discover wiring) matches the issue's stated ask. As a static markdown addition with all CI checks green, there's nothing here that can break the build or tests.

Nits — 6 non-blocking

• The claim '99/100' described labels for rust-lang/rust (a repo the doc says has '100+' labels) implies the count was capped at the first page or a specific snapshot — worth a footnote clarifying whether this is exact or truncated by pagination.
• The doc states figures are from 2026-07-17 but gives no raw data dump or links to the underlying API responses, so the numbers are unverifiable by a future reader without re-running the audit themselves.
• Table row for `sindresorhus/slugify` lists 8 labels 'all described' but the negative-case section right below cites `sindresorhus/p-map` (7 labels) and `chalk/chalk` (10 labels) as supporting examples that aren't in the main sample table, which could confuse a reader about the actual sample size (10 vs. 12 repos referenced).
• Consider adding an appendix with raw JSON snippets or a script for reproducibility, per the 'Method / reproducibility' section's own emphasis on this being derived from live data, not memory.
• Clarify in the sample table or a footnote whether `p-map` and `chalk` are supplementary spot-checks or part of the core 10, since they're referenced later but not tabulated.
• Diff looks like trivial or whitespace-only churn — Reduce whitespace-only or formatting-only churn and keep the diff focused on substantive changes.

Decision drivers

• ✅ Code review — No blockers (1 reviewer)
• ✅ Gate result — Passing (No configured blocker found.)

Context & advisory signals — never blocks the verdict

Signal	Result	Evidence
Linked issue	✅ Linked	#6794
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	✅ 20/20	Low review scope from cached public metadata (1 linked issue).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 172 registered-repo PR(s), 103 merged, 27 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	luciferlive112116; Gittensor profile; 172 PR(s), 27 issue(s).
Improvement	ℹ️ Insufficient signal	risk: low · value: insufficient-signal

Review context

• Author: luciferlive112116
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: not available
• Official Gittensor activity: 172 PR(s), 27 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Keep the PR focused and include validation evidence before maintainer review.

Signal definitions

• Related work = same linked issue, overlapping active PRs, or title/path similarity.
• Change scope = cached public metadata such as size labels, draft state, and review-burden hints.
• Validation posture = whether the PR provides enough public validation/test evidence for maintainer review.
• Contributor workload = public contributor activity and cleanup pressure, not a repo-wide quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

🧪 Chat with LoopOver

Ask LoopOver a question about this PR directly in a comment — grounded only in the same cached, public-safe facts shown above, never a new claim.

• @loopover ask <question> answers contribution-quality Q&A with source citations and freshness.
• @loopover chat <question> answers in natural prose from cached decision-pack facts via local inference (maintainer/collaborator; read-only).
• A plain-language @loopover mention with a real question is routed to the closest matching read-only command automatically — no exact syntax required.

Full command reference: https://loopover.ai/docs/loopover-commands

🧪 Experimental — new and may change.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

---

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by LoopOver, a quiet PR intelligence layer for OSS maintainers.

• Re-run LoopOver review