erdos-problems 0.1.7 → 0.1.8

package/problems/2/problem.yaml

@@ -0,0 +1,42 @@
+problem_id: "2"
+display_name: "Erdos Problem #2"
+title: "Erdos Problem #2"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/2
+  external_id: "2"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "2"
+provenance:
+  seeded_at: 2026-03-26T04:41:25.055Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: disproved
+  site_badge: DISPROVED
+  repo_status: cataloged
+  upstream_status: disproved
+  upstream_last_update: 2025-08-31
+cluster: number-theory
+prize:
+  display: $1000
+related_problems: []
+family_tags:
+  - number-theory
+  - covering-systems
+harness:
+  depth: dossier
+statement:
+  short: "Curated dossier seeded from upstream public metadata for Erdos Problem #2."
+  normalized_md_path: STATEMENT.md
+references_path: REFERENCES.md
+evidence_path: EVIDENCE.md
+formalization_path: FORMALIZATION.md
+formalization:
+  status: planned
+  upstream_state: no
+  upstream_last_update: 2025-08-31

package/problems/21/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 21`
+- `erdos workspace show`
+- `erdos preflight`
+- `erdos continuation use route`
+- `erdos checkpoints sync`
+
+Working assumptions:
+- Open problem: yes
+- Active route: seed_route_pending
+- Repo status: cataloged
+- Harness depth: dossier
+
+First honest move:
+- tighten the local dossier for problem 21 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/21/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 21
+- Repo status: cataloged
+- Harness depth: dossier
+
+Checkpoint prompts:
+- What changed in the active route since the last honest checkpoint?
+- Which claim level is justified right now: Exact, Verified, Heuristic, or Conjecture?
+- Which upstream/public truth changed, if any?
+- Which artifact or literature bundle should the next agent read first?

package/problems/21/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #21 from a pull bundle.
+- Upstream record included: yes
+- Site snapshot included: yes
+- Repo status at seed time: cataloged
+- Harness depth at seed time: dossier
+
+Next maintainer step:
+- replace the seed placeholders with a curated statement, references, evidence notes, and formalization plan.

package/problems/21/FORMALIZATION.md

@@ -0,0 +1,8 @@
+# Formalization
+
+- Local status: planned
+- Upstream formalized state: no
+- Upstream formalized last update: 2025-08-31
+
+Seed note:
+- this file was created automatically from a pull bundle and should be upgraded as local formal work begins.

package/problems/21/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/21>
+- Upstream structured data:
+  - <https://github.com/teorth/erdosproblems>
+  - `data/problems.yaml`
+- Site summary snapshot was present in the pull bundle at seed time.
+- A literature lane bundle was available during seeding.

package/problems/21/ROUTES.md

@@ -0,0 +1,14 @@
+# Routes
+
+## Status Ladder
+
+- Open problem: yes
+- Active route: seed_route_pending
+- Route breakthrough: no
+- Problem solved: no
+
+## Starter route notes
+
+- Current seeded route placeholder for problem 21: `seed_route_pending`
+- Treat this as a workspace-level route marker until a curated route tree is written.
+- Keep route progress separate from global problem status.

package/problems/21/STATEMENT.md

@@ -0,0 +1,21 @@
+# Problem 21 Statement
+
+Source: <https://www.erdosproblems.com/21>
+
+Normalized focus:
+- Let $f(n)$ be minimal such that there is an intersecting family $\mathcal{F}$ of sets of size $n$ (so $A\cap B\neq\emptyset$ for all $A,B\in \mathcal{F}$) with $\lvert \mathcal{F}\rvert=f(n)$ such that any set $S$ with $\lvert S\rvert \leq n-1$ is disjoint from at least one $A\in \mathcal{F}$.
+- Seeded with filtered statement candidates from the public site snapshot
+
+Statement candidates:
+- Let $f(n)$ be minimal such that there is an intersecting family $\mathcal{F}$ of sets of size $n$ (so $A\cap B\neq\emptyset$ for all $A,B\in \mathcal{F}$) with $\lvert \mathcal{F}\rvert=f(n)$ such that any set $S$ with $\lvert S\rvert \leq n-1$ is disjoint from at least one $A\in \mathcal{F}$.
+- Is it true that\[f(n) \ll n?\]
+- Conjectured by Erdős and Lovász [ErLo75] , who proved that\[\frac{8}{3}n-3\leq f(n) \ll n^{3/2}\log n\]for all $n$. The upper bound was improved by Kahn [Ka92b] to\[f(n) \ll n\log n.\](The upper bound constructions in both cases are formed by taking a random set of lines from a projective plane of order $n-1$, assuming $n-1$ is a prime power.)
+- This problem was solved by Kahn [Ka94] who proved the upper bound $f(n) \ll n$. The Erdős-Lovász lower bound of $\frac{8}{3}n-O(1)$ has not been improved, and it has been speculated (see e.g. [Ka94] ) that the correct answer is $3n+O(1)$.
+
+Public-site preview:
+- PROVED
+- This has been solved in the affirmative.
+- - $500
+- Let $f(n)$ be minimal such that there is an intersecting family $\mathcal{F}$ of sets of size $n$ (so $A\cap B\neq\emptyset$ for all $A,B\in \mathcal{F}$) with $\lvert \mathcal{F}\rvert=f(n)$ such that any set $S$ with $\lvert S\rvert \leq n-1$ is disjoint from at least one $A\in \mathcal{F}$.
+- Is it true that\[f(n) \ll n?\]
+- #21 : [Er81] [Er90] [Er92b] [Er97f]

package/problems/21/problem.yaml

@@ -0,0 +1,46 @@
+problem_id: "21"
+display_name: "Erdos Problem #21"
+title: "Erdos Problem #21"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/21
+  external_id: "21"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "21"
+provenance:
+  seeded_at: 2026-03-26T04:41:26.922Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: proved
+  site_badge: PROVED
+  repo_status: cataloged
+  upstream_status: proved
+  upstream_last_update: 2025-08-31
+cluster: combinatorics
+prize:
+  display: $500
+related_problems: []
+family_tags:
+  - combinatorics
+  - intersecting-family
+harness:
+  depth: dossier
+statement:
+  short: Let $f(n)$ be minimal such that there is an intersecting family
+    $\mathcal{F}$ of sets of size $n$ (so $A\cap B\neq\emptyset$ for all $A,B\in
+    \mathcal{F}$) with $\lvert \mathcal{F}\rvert=f(n)$ such that any set $S$
+    with $\lvert S\rvert \leq n-1$ is disjoint from at least one $A\in
+    \mathcal{F}$.
+  normalized_md_path: STATEMENT.md
+references_path: REFERENCES.md
+evidence_path: EVIDENCE.md
+formalization_path: FORMALIZATION.md
+formalization:
+  status: planned
+  upstream_state: no
+  upstream_last_update: 2025-08-31

package/problems/22/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 22`
+- `erdos workspace show`
+- `erdos preflight`
+- `erdos continuation use route`
+- `erdos checkpoints sync`
+
+Working assumptions:
+- Open problem: yes
+- Active route: seed_route_pending
+- Repo status: cataloged
+- Harness depth: dossier
+
+First honest move:
+- tighten the local dossier for problem 22 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/22/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 22
+- Repo status: cataloged
+- Harness depth: dossier
+
+Checkpoint prompts:
+- What changed in the active route since the last honest checkpoint?
+- Which claim level is justified right now: Exact, Verified, Heuristic, or Conjecture?
+- Which upstream/public truth changed, if any?
+- Which artifact or literature bundle should the next agent read first?

package/problems/22/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #22 from a pull bundle.
+- Upstream record included: yes
+- Site snapshot included: yes
+- Repo status at seed time: cataloged
+- Harness depth at seed time: dossier
+
+Next maintainer step:
+- replace the seed placeholders with a curated statement, references, evidence notes, and formalization plan.

package/problems/22/FORMALIZATION.md

@@ -0,0 +1,8 @@
+# Formalization
+
+- Local status: planned
+- Upstream formalized state: no
+- Upstream formalized last update: 2025-08-31
+
+Seed note:
+- this file was created automatically from a pull bundle and should be upgraded as local formal work begins.

package/problems/22/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/22>
+- Upstream structured data:
+  - <https://github.com/teorth/erdosproblems>
+  - `data/problems.yaml`
+- Site summary snapshot was present in the pull bundle at seed time.
+- A literature lane bundle was available during seeding.

package/problems/22/ROUTES.md

@@ -0,0 +1,14 @@
+# Routes
+
+## Status Ladder
+
+- Open problem: yes
+- Active route: seed_route_pending
+- Route breakthrough: no
+- Problem solved: no
+
+## Starter route notes
+
+- Current seeded route placeholder for problem 22: `seed_route_pending`
+- Treat this as a workspace-level route marker until a curated route tree is written.
+- Keep route progress separate from global problem status.

package/problems/22/STATEMENT.md

@@ -0,0 +1,21 @@
+# Problem 22 Statement
+
+Source: <https://www.erdosproblems.com/22>
+
+Normalized focus:
+- Let $\epsilon>0$ and $n$ be sufficiently large depending on $\epsilon$. Is there a graph on $n$ vertices with $\geq n^2/8$ many edges which contains no $K_4$ such that the largest independent set has size at most $\epsilon n$?
+- Seeded with filtered statement candidates from the public site snapshot
+
+Statement candidates:
+- Let $\epsilon>0$ and $n$ be sufficiently large depending on $\epsilon$. Is there a graph on $n$ vertices with $\geq n^2/8$ many edges which contains no $K_4$ such that the largest independent set has size at most $\epsilon n$?
+- In other words, if $\mathrm{rt}(n;k,\ell)$ is the Ramsey-Turán number then is it true that (for sufficiently large $n$)\[\mathrm{rt}(n; 4,\epsilon n)\geq n^2/8?\]Conjectured by Bollobás and Erdős [BoEr76] , who proved the existence of such a graph with $(1/8+o(1))n^2$ many edges. Solved by Fox, Loh, and Zhao [FLZ15] , who proved that for every $n\geq 1$ there exists a graph on $n$ vertices with $\geq n^2/8$ many edges, containing no $K_4$, whose largest independent set has size at most\[ \ll \frac{(\log\log n)^{3/2}}{(\log n)^{1/2}}n.\]See also [615] .
+- No ( Create a formalisation here )
+- Related OEIS sequences:
+
+Public-site preview:
+- PROVED
+- This has been solved in the affirmative.
+- Let $\epsilon>0$ and $n$ be sufficiently large depending on $\epsilon$. Is there a graph on $n$ vertices with $\geq n^2/8$ many edges which contains no $K_4$ such that the largest independent set has size at most $\epsilon n$?
+- #22 : [BoEr76] [Er90]
+- graph theory
+- In other words, if $\mathrm{rt}(n;k,\ell)$ is the Ramsey-Turán number then is it true that (for sufficiently large $n$)\[\mathrm{rt}(n; 4,\epsilon n)\geq n^2/8?\]Conjectured by Bollobás and Erdős [BoEr76] , who proved the existence of such a graph with $(1/8+o(1))n^2$ many edges. Solved by Fox, Loh, and Zhao [FLZ15] , who proved that for every $n\geq 1$ there exists a graph on $n$ vertices with $\geq n^2/8$ many edges, containing no $K_4$, whose largest independent set has size at most\[ \ll \frac{(\log\log n)^{3/2}}{(\log n)^{1/2}}n.\]See also [615] .

package/problems/22/problem.yaml

@@ -0,0 +1,44 @@
+problem_id: "22"
+display_name: "Erdos Problem #22"
+title: "Erdos Problem #22"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/22
+  external_id: "22"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "22"
+provenance:
+  seeded_at: 2026-03-26T04:41:27.183Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: proved
+  site_badge: PROVED
+  repo_status: cataloged
+  upstream_status: proved
+  upstream_last_update: 2025-08-31
+cluster: graph-theory
+prize:
+  display: no
+related_problems: []
+family_tags:
+  - graph-theory
+harness:
+  depth: dossier
+statement:
+  short: Let $\epsilon>0$ and $n$ be sufficiently large depending on $\epsilon$.
+    Is there a graph on $n$ vertices with $\geq n^2/8$ many edges which contains
+    no $K_4$ such that the largest independent set has size at most $\epsilon
+    n$?
+  normalized_md_path: STATEMENT.md
+references_path: REFERENCES.md
+evidence_path: EVIDENCE.md
+formalization_path: FORMALIZATION.md
+formalization:
+  status: planned
+  upstream_state: no
+  upstream_last_update: 2025-08-31

package/problems/3/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 3`
+- `erdos workspace show`
+- `erdos preflight`
+- `erdos continuation use route`
+- `erdos checkpoints sync`
+
+Working assumptions:
+- Open problem: yes
+- Active route: seed_route_pending
+- Repo status: cataloged
+- Harness depth: dossier
+
+First honest move:
+- tighten the local dossier for problem 3 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/3/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 3
+- Repo status: cataloged
+- Harness depth: dossier
+
+Checkpoint prompts:
+- What changed in the active route since the last honest checkpoint?
+- Which claim level is justified right now: Exact, Verified, Heuristic, or Conjecture?
+- Which upstream/public truth changed, if any?
+- Which artifact or literature bundle should the next agent read first?

package/problems/3/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #3 from a pull bundle.
+- Upstream record included: yes
+- Site snapshot included: yes
+- Repo status at seed time: cataloged
+- Harness depth at seed time: dossier
+
+Next maintainer step:
+- replace the seed placeholders with a curated statement, references, evidence notes, and formalization plan.

package/problems/3/FORMALIZATION.md

@@ -0,0 +1,8 @@
+# Formalization
+
+- Local status: planned
+- Upstream formalized state: yes
+- Upstream formalized last update: 2025-08-31
+
+Seed note:
+- this file was created automatically from a pull bundle and should be upgraded as local formal work begins.

package/problems/3/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/3>
+- Upstream structured data:
+  - <https://github.com/teorth/erdosproblems>
+  - `data/problems.yaml`
+- Site summary snapshot was present in the pull bundle at seed time.
+- A literature lane bundle was available during seeding.

package/problems/3/ROUTES.md

@@ -0,0 +1,14 @@
+# Routes
+
+## Status Ladder
+
+- Open problem: yes
+- Active route: seed_route_pending
+- Route breakthrough: no
+- Problem solved: no
+
+## Starter route notes
+
+- Current seeded route placeholder for problem 3: `seed_route_pending`
+- Treat this as a workspace-level route marker until a curated route tree is written.
+- Keep route progress separate from global problem status.

package/problems/3/STATEMENT.md

@@ -0,0 +1,21 @@
+# Problem 3 Statement
+
+Source: <https://www.erdosproblems.com/3>
+
+Normalized focus:
+- If $A\subseteq \mathbb{N}$ has $\sum_{n\in A}\frac{1}{n}=\infty$ then must $A$ contain arbitrarily long arithmetic progressions?
+- Seeded with filtered statement candidates from the public site snapshot
+
+Statement candidates:
+- If $A\subseteq \mathbb{N}$ has $\sum_{n\in A}\frac{1}{n}=\infty$ then must $A$ contain arbitrarily long arithmetic progressions?
+- This is essentially asking for good bounds on $r_k(N)$, the size of the largest subset of $\{1,\ldots,N\}$ without a non-trivial $k$-term arithmetic progression. For example, a bound like\[r_k(N) \ll_k \frac{N}{(\log N)(\log\log N)^2}\]would be sufficient.
+- Even the case $k=3$ is non-trivial, but was proved by Bloom and Sisask [BlSi20] . Much better bounds for $r_3(N)$ were subsequently proved by Kelley and Meka [KeMe23] . Green and Tao [GrTa17] proved $r_4(N)\ll N/(\log N)^{c}$ for some small constant $c>0$. Gowers [Go01] proved\[r_k(N) \ll \frac{N}{(\log\log N)^{c_k}},\]where $c_k>0$ is a small constant depending on $k$. The current best bounds for general $k$ are due to Leng, Sah, and Sawhney [LSS24] , who show that\[r_k(N) \ll \frac{N}{\exp((\log\log N)^{c_k})}\]for some constant $c_k>0$ depending on $k$.
+- Curiously, Erdős [Er83c] thought this conjecture was the 'only way to approach' the conjecture that there are arbitrarily long arithmetic progressions of prime numbers, now a theorem due to Green and Tao [GrTa08] (see [219] ).
+
+Public-site preview:
+- OPEN
+- This is open, and cannot be resolved with a finite computation.
+- - $5000
+- If $A\subseteq \mathbb{N}$ has $\sum_{n\in A}\frac{1}{n}=\infty$ then must $A$ contain arbitrarily long arithmetic progressions?
+- #3 : [Er74b] [Er75b] [Er77c] [ErGr79] [Er80] [Er80c] [ErGr80,p.11] [Er81] [Er82e] [Er83] [Er83c] [Er85c] [Er90] [Er97c] [Va99,1.28]
+- number theory |

package/problems/3/problem.yaml

@@ -0,0 +1,44 @@
+problem_id: "3"
+display_name: "Erdos Problem #3"
+title: "Erdos Problem #3"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/3
+  external_id: "3"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "3"
+provenance:
+  seeded_at: 2026-03-26T04:41:25.280Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: open
+  site_badge: OPEN
+  repo_status: cataloged
+  upstream_status: open
+  upstream_last_update: 2025-08-31
+cluster: number-theory
+prize:
+  display: $5000
+related_problems: []
+family_tags:
+  - number-theory
+  - additive-combinatorics
+  - arithmetic-progressions
+harness:
+  depth: dossier
+statement:
+  short: If $A\subseteq \mathbb{N}$ has $\sum_{n\in A}\frac{1}{n}=\infty$ then
+    must $A$ contain arbitrarily long arithmetic progressions?
+  normalized_md_path: STATEMENT.md
+references_path: REFERENCES.md
+evidence_path: EVIDENCE.md
+formalization_path: FORMALIZATION.md
+formalization:
+  status: planned
+  upstream_state: yes
+  upstream_last_update: 2025-08-31

package/problems/4/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 4`
+- `erdos workspace show`
+- `erdos preflight`
+- `erdos continuation use route`
+- `erdos checkpoints sync`
+
+Working assumptions:
+- Open problem: yes
+- Active route: seed_route_pending
+- Repo status: cataloged
+- Harness depth: dossier
+
+First honest move:
+- tighten the local dossier for problem 4 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/4/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 4
+- Repo status: cataloged
+- Harness depth: dossier
+
+Checkpoint prompts:
+- What changed in the active route since the last honest checkpoint?
+- Which claim level is justified right now: Exact, Verified, Heuristic, or Conjecture?
+- Which upstream/public truth changed, if any?
+- Which artifact or literature bundle should the next agent read first?

package/problems/4/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #4 from a pull bundle.
+- Upstream record included: yes
+- Site snapshot included: yes
+- Repo status at seed time: cataloged
+- Harness depth at seed time: dossier
+
+Next maintainer step:
+- replace the seed placeholders with a curated statement, references, evidence notes, and formalization plan.

package/problems/4/FORMALIZATION.md

@@ -0,0 +1,8 @@
+# Formalization
+
+- Local status: planned
+- Upstream formalized state: yes
+- Upstream formalized last update: 2025-08-31
+
+Seed note:
+- this file was created automatically from a pull bundle and should be upgraded as local formal work begins.

package/problems/4/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/4>
+- Upstream structured data:
+  - <https://github.com/teorth/erdosproblems>
+  - `data/problems.yaml`
+- Site summary snapshot was present in the pull bundle at seed time.
+- A literature lane bundle was available during seeding.

package/problems/4/ROUTES.md

@@ -0,0 +1,14 @@
+# Routes
+
+## Status Ladder
+
+- Open problem: yes
+- Active route: seed_route_pending
+- Route breakthrough: no
+- Problem solved: no
+
+## Starter route notes
+
+- Current seeded route placeholder for problem 4: `seed_route_pending`
+- Treat this as a workspace-level route marker until a curated route tree is written.
+- Keep route progress separate from global problem status.

package/problems/4/STATEMENT.md

@@ -0,0 +1,21 @@
+# Problem 4 Statement
+
+Source: <https://www.erdosproblems.com/4>
+
+Normalized focus:
+- Is it true that, for any $C>0$, there are infinitely many $n$ such that\[p_{n+1}-p_n> C\frac{\log\log n\log\log\log\log n}{(\log\log \log n)^2}\log n?\]
+- Seeded with filtered statement candidates from the public site snapshot
+
+Statement candidates:
+- Is it true that, for any $C>0$, there are infinitely many $n$ such that\[p_{n+1}-p_n> C\frac{\log\log n\log\log\log\log n}{(\log\log \log n)^2}\log n?\]
+- The peculiar quantitative form of Erdős' question was motivated by an old result of Rankin [Ra38] , who proved there exists some constant $C>0$ such that the claim holds. Solved by Maynard [Ma16] and Ford, Green, Konyagin, and Tao [FGKT16] . The best bound available, due to all five authors [FGKMT18] , is that there are infinitely many $n$ such that\[p_{n+1}-p_n\gg \frac{\log\log n\log\log\log\log n}{\log\log \log n}\log n.\]The likely truth is a lower bound like $\gg(\log n)^2$. In [Er97c] Erdős revised the value of this problem to \$5000 and reserved the \$10000 for a lower bound of $>(\log n)^{1+c}$ for some $c>0$.
+- The best known upper bound is\[p_{n+1}-p_n \ll n^{0.525+o(1)},\]proved by Baker, Harman, and Pintz [BHP01] .
+- See also [687] .
+
+Public-site preview:
+- PROVED
+- This has been solved in the affirmative.
+- - $10000
+- Is it true that, for any $C>0$, there are infinitely many $n$ such that\[p_{n+1}-p_n> C\frac{\log\log n\log\log\log\log n}{(\log\log \log n)^2}\log n?\]
+- #4 : [Er55c,p.2] [Er57,p.292] [Er61,p.223] [Er65b,p.201] [Er81k,p.1] [Er82e,p.64] [Er90] [Er97c] [Er97f] [Va99,1.1]
+- number theory |

package/problems/4/problem.yaml

@@ -0,0 +1,44 @@
+problem_id: "4"
+display_name: "Erdos Problem #4"
+title: "Erdos Problem #4"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/4
+  external_id: "4"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "4"
+provenance:
+  seeded_at: 2026-03-26T04:41:25.574Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: proved
+  site_badge: PROVED
+  repo_status: cataloged
+  upstream_status: proved
+  upstream_last_update: 2025-08-31
+cluster: number-theory
+prize:
+  display: $10000
+related_problems: []
+family_tags:
+  - number-theory
+  - primes
+harness:
+  depth: dossier
+statement:
+  short: Is it true that, for any $C>0$, there are infinitely many $n$ such
+    that\[p_{n+1}-p_n> C\frac{\log\log n\log\log\log\log n}{(\log\log \log
+    n)^2}\log n?\]
+  normalized_md_path: STATEMENT.md
+references_path: REFERENCES.md
+evidence_path: EVIDENCE.md
+formalization_path: FORMALIZATION.md
+formalization:
+  status: planned
+  upstream_state: yes
+  upstream_last_update: 2025-08-31

package/problems/5/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 5`
+- `erdos workspace show`
+- `erdos preflight`
+- `erdos continuation use route`
+- `erdos checkpoints sync`
+
+Working assumptions:
+- Open problem: yes
+- Active route: seed_route_pending
+- Repo status: cataloged
+- Harness depth: dossier
+
+First honest move:
+- tighten the local dossier for problem 5 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/5/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 5
+- Repo status: cataloged
+- Harness depth: dossier
+
+Checkpoint prompts:
+- What changed in the active route since the last honest checkpoint?
+- Which claim level is justified right now: Exact, Verified, Heuristic, or Conjecture?
+- Which upstream/public truth changed, if any?
+- Which artifact or literature bundle should the next agent read first?