erdos-problems 0.1.6 → 0.1.8

package/README.md

@@ -26,7 +26,10 @@ Official binary:
 - unseeded problems can still be pulled into a workspace from the bundled upstream snapshot
 
 Seeded problems:
-- `18`, `20`, `89`, `536`, `542`, `856`, `857`, `1008`
+- `1`, `2`, `3`, `4`, `5`, `6`, `7`, `18`, `19`, `20`, `21`, `22`, `89`, `536`, `542`, `856`, `857`, `1008`
+
+Native dossier count:
+- `18`
 
 ## First-run flow
 
@@ -38,6 +41,14 @@ erdos sunflower status 857
 erdos dossier show 857
 ```
 
+For an unseeded problem, the one-step self-seeding flow is now:
+
+```bash
+erdos seed problem 25 --include-site --cluster number-theory
+erdos problem show 25
+erdos workspace show
+```
+
 What `bootstrap` does:
 - sets the active workspace problem
 - scaffolds the canonical dossier files into `.erdos/scaffolds/<id>/`
@@ -45,6 +56,17 @@ What `bootstrap` does:
 - copies pack-specific context and compute packets when the problem has them
 - gives an agent a ready-to-read local artifact bundle immediately after install
 
+What `seed` does:
+- creates a pull bundle for any problem in the upstream snapshot
+- promotes that bundle into `.erdos/seeded-problems/<id>/`
+- auto-selects the problem in the workspace
+- syncs the staged research loop state and checkpoint shelf
+- makes the new dossier visible to the atlas commands immediately inside that workspace
+- writes richer starter artifacts:
+  - `AGENT_START.md`
+  - `ROUTES.md`
+  - `CHECKPOINT_NOTES.md`
+
 ## Pull lanes
 
 For any problem number in the upstream snapshot, you can create a workspace bundle even if the problem is not yet seeded locally:
@@ -72,10 +94,9 @@ What the pull lanes do:
 To turn a pulled bundle into a new canonical dossier in the repo:
 
 ```bash
-erdos maintainer seed problem 1 \
-  --from-pull .erdos/pulls/1 \
-  --cluster number-theory \
-  --family-tag additive-combinatorics
+erdos maintainer seed problem 25 \
+  --from-pull .erdos/pulls/25 \
+  --cluster number-theory
 ```
 
 What maintainer seeding does:
@@ -87,6 +108,10 @@ What maintainer seeding does:
   - `REFERENCES.md`
   - `EVIDENCE.md`
   - `FORMALIZATION.md`
+- starter loop files:
+  - `AGENT_START.md`
+  - `ROUTES.md`
+  - `CHECKPOINT_NOTES.md`
 - preserves upstream/site provenance in the local record
 
 ## Sunflower pack
@@ -97,6 +122,10 @@ The first deep pack is the sunflower quartet:
 - `536`: natural-density LCM analogue
 - `856`: harmonic-density LCM analogue
 
+Deep sunflower problems now also ship route packets:
+- `20`: `AGENT_START.md`, `ROUTE_PACKET.yaml`, `CHECKPOINT_PACKET.md`, `REPORT_PACKET.md`
+- `857`: `AGENT_START.md`, `ROUTE_PACKET.yaml`, `CHECKPOINT_PACKET.md`, `REPORT_PACKET.md`
+
 Useful sunflower commands:
 
 ```bash
@@ -138,10 +167,11 @@ erdos upstream diff
 erdos scaffold problem 857
 erdos bootstrap problem 857
 erdos bootstrap problem 857 --sync-upstream
+erdos seed problem 25 --include-site --cluster number-theory
 erdos pull problem 857
 erdos pull artifacts 857
 erdos pull literature 857
-erdos maintainer seed problem 1 --from-pull .erdos/pulls/1 --cluster number-theory
+erdos maintainer seed problem 25 --from-pull .erdos/pulls/25 --cluster number-theory
 ```
 
 ## Canonical sources
@@ -161,17 +191,24 @@ For each seeded problem, the canonical local dossier lives in `problems/<id>/`:
 - `EVIDENCE.md`
 - `FORMALIZATION.md`
 
+Many seeded dossiers now also carry starter-loop artifacts:
+- `AGENT_START.md`
+- `ROUTES.md`
+- `CHECKPOINT_NOTES.md`
+
 The CLI can surface these directly:
 - `erdos problem artifacts <id>` shows the canonical inventory
 - `erdos problem artifacts <id> --json` emits machine-readable inventory
 - `erdos scaffold problem <id>` copies the seeded dossier into the active workspace
 - `erdos bootstrap problem <id>` selects the problem and creates the scaffold in one step
+- `erdos seed problem <id>` self-seeds an unseeded problem into `.erdos/seeded-problems/` and syncs the loop
 - `erdos pull problem <id>` creates a workspace bundle for any problem in the upstream snapshot
 - `erdos maintainer seed problem <id>` promotes a pull bundle into a canonical local dossier
 
 For sunflower problems, the CLI also surfaces pack-specific artifacts:
 - pack README context
 - per-problem context files under `packs/sunflower/problems/<id>/`
+- route packets and checkpoint/report packets for `20` and `857`
 - compute packets under `packs/sunflower/compute/<id>/` when available
 
 ## Notes
@@ -213,5 +250,9 @@ The public package uses the same status ladder we settled on in the lab:
 - route breakthrough
 - problem solved
 
+That means a fresh install now supports two clean research starts:
+- `erdos bootstrap problem <seeded-id>` for native packaged dossiers
+- `erdos seed problem <unseeded-id>` for workspace-local self-seeding with the same loop
+
 See also:
 - `docs/RESEARCH_LOOP.md`

package/docs/RESEARCH_LOOP.md

@@ -36,6 +36,22 @@ erdos checkpoints sync
 erdos workspace show
 ```
 
+For problems that are not yet packaged as native dossiers, the loop can start with one-step self-seeding:
+
+```bash
+erdos seed problem 25 --include-site --cluster number-theory
+erdos preflight
+erdos continuation use route
+erdos checkpoints sync
+```
+
+That flow writes the local dossier into `.erdos/seeded-problems/<id>/`, makes it visible to the atlas inside the current workspace, and immediately syncs the same state/checkpoint machinery used by packaged dossiers.
+
+The seeded dossier now also includes starter-loop artifacts:
+- `AGENT_START.md`
+- `ROUTES.md`
+- `CHECKPOINT_NOTES.md`
+
 ## Status ladder
 
 The public package uses the same ladder we converged on in the lab:

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "erdos-problems",
-  "version": "0.1.6",
+  "version": "0.1.8",
   "description": "CLI atlas and staged research harness for Paul Erdos problems.",
   "type": "module",
   "bin": {

package/packs/sunflower/README.md

@@ -13,3 +13,7 @@ Profiles:
 - `857`: deep harness, weak / non-uniform core
 - `536`: dossier bridge, natural-density LCM analogue
 - `856`: dossier bridge, harmonic-density LCM analogue
+
+Deep route packets:
+- `20` ships `AGENT_START.md`, `ROUTE_PACKET.yaml`, `CHECKPOINT_PACKET.md`, `REPORT_PACKET.md`
+- `857` ships `AGENT_START.md`, `ROUTE_PACKET.yaml`, `CHECKPOINT_PACKET.md`, `REPORT_PACKET.md`

package/packs/sunflower/problems/20/AGENT_START.md

@@ -0,0 +1,18 @@
+# Problem 20 Agent Start
+
+Start here:
+- `erdos problem show 20`
+- `erdos sunflower status 20`
+- `erdos preflight`
+- `erdos checkpoints sync`
+
+Current framing:
+- open problem: yes
+- active route: `uniform_k3_frontier`
+- route breakthrough: no
+- problem solved: no
+
+Keep the work honest:
+- stay in the uniform-family language
+- preserve every bridge back to `857`
+- checkpoint any route breakthrough before widening the claim

package/packs/sunflower/problems/20/CHECKPOINT_PACKET.md

@@ -0,0 +1,11 @@
+# Problem 20 Checkpoint Packet
+
+Checkpoint focus:
+- keep the active route stated in uniform-family language
+- record any bridge claim to `857` explicitly
+- separate route progress from full strong-sunflower closure
+
+Questions:
+- what changed in the active uniform k=3 route?
+- what is now reusable outside the immediate local proof step?
+- what still blocks recurrence-facing transfer?

package/packs/sunflower/problems/20/REPORT_PACKET.md

@@ -0,0 +1,10 @@
+# Problem 20 Report Packet
+
+Suggested report sections:
+- active route snapshot
+- current frontier obligation
+- bridge notes to `857`
+- verification hook
+- next honest move
+
+Do not collapse local route progress into a claim that the strong sunflower problem is solved.

package/packs/sunflower/problems/20/ROUTE_PACKET.yaml

@@ -0,0 +1,12 @@
+route_packet_id: strong20_uniform_k3_frontier_v1
+route_id: uniform_k3_frontier
+frontier_claim: Tighten the uniform k=3 sunflower lane into a recurrence-ready and transfer-ready packet that can honestly talk to problem 857.
+theorem_module: sunflower-coda/repo/sunflower_lean/SunflowerLean/ErdosProblem20.lean
+checkpoint_packet: CHECKPOINT_PACKET.md
+report_packet: REPORT_PACKET.md
+ready_prompts:
+  - Which uniform k=3 reduction is the next smallest honest step?
+  - Which reusable bridge would transfer directly back into the weak 857 program?
+verification_hook:
+  - erdos sunflower status 20
+  - erdos problem artifacts 20 --json

package/packs/sunflower/problems/857/AGENT_START.md

@@ -0,0 +1,18 @@
+# Problem 857 Agent Start
+
+Start here:
+- `erdos problem show 857`
+- `erdos sunflower status 857`
+- `erdos preflight`
+- `erdos checkpoints sync`
+
+Current framing:
+- open problem: yes
+- active route: `anchored_selector_linearization`
+- route breakthrough: yes
+- problem solved: no
+
+Keep the work honest:
+- preserve the distinction between active-route progress and full problem closure
+- keep the recurrence-facing remainder terms explicit
+- checkpoint every route breakthrough before selecting the next route

package/packs/sunflower/problems/857/CHECKPOINT_PACKET.md

@@ -0,0 +1,11 @@
+# Problem 857 Checkpoint Packet
+
+Checkpoint focus:
+- record the active route and the exact remainder/compression seam
+- preserve open-problem versus active-route language
+- keep reusable export and recurrence notes separate from speculative closure claims
+
+Questions:
+- what changed in the explicit O1a export package?
+- which remainder term is still dominating?
+- what is the next honest compression step?

package/packs/sunflower/problems/857/REPORT_PACKET.md

@@ -0,0 +1,10 @@
+# Problem 857 Report Packet
+
+Suggested report sections:
+- active route snapshot
+- current export/compression frontier
+- explicit remainder terms still alive
+- verification hook
+- next honest move
+
+Do not phrase a route breakthrough as a solved weak sunflower problem.

package/packs/sunflower/problems/857/ROUTE_PACKET.yaml

@@ -0,0 +1,12 @@
+route_packet_id: weak857_export_compression_v1
+route_id: anchored_selector_linearization
+frontier_claim: Preserve the live export and compression frontier and keep pushing the explicit O1a remainder package toward a genuine recurrence-facing step.
+theorem_module: sunflower-coda/repo/sunflower_lean/SunflowerLean/ObstructionGlobalBridge.lean
+checkpoint_packet: CHECKPOINT_PACKET.md
+report_packet: REPORT_PACKET.md
+ready_prompts:
+  - Which explicit compression step is the next honest reduction on the active weak-sunflower route?
+  - Which exported remainder term is still dominating the recurrence-facing frontier?
+verification_hook:
+  - erdos sunflower status 857
+  - erdos problem artifacts 857 --json

package/problems/1/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 1`
+- `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 1 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/1/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 1
+- 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/1/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #1 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/1/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/1/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/1>
+- 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/1/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 1: `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/1/STATEMENT.md

@@ -0,0 +1,21 @@
+# Problem 1 Statement
+
+Source: <https://www.erdosproblems.com/1>
+
+Normalized focus:
+- If $A\subseteq \{1,\ldots,N\}$ with $\lvert A\rvert=n$ is such that the subset sums $\sum_{a\in S}a$ are distinct for all $S\subseteq A$ then\[N \gg 2^{n}.\]
+- Seeded with filtered statement candidates from the public site snapshot
+
+Statement candidates:
+- If $A\subseteq \{1,\ldots,N\}$ with $\lvert A\rvert=n$ is such that the subset sums $\sum_{a\in S}a$ are distinct for all $S\subseteq A$ then\[N \gg 2^{n}.\]
+- Erdős called this 'perhaps my first serious problem' (in [Er98] he dates it to 1931). The powers of $2$ show that $2^n$ would be best possible here. The trivial lower bound is $N \gg 2^{n}/n$, since all $2^n$ distinct subset sums must lie in $[0,Nn)$. Erdős and Moser [Er56] proved\[ N\geq (\tfrac{1}{4}-o(1))\frac{2^n}{\sqrt{n}}.\](In [Er85c] Erdős offered \$100 for any improvement of the constant $1/4$ here.)
+- A number of improvements of the constant have been given (see [St23] for a history), with the current record $\sqrt{2/\pi}$ first proved in unpublished work of Elkies and Gleason. Two proofs achieving this constant are provided by Dubroff, Fox, and Xu [DFX21] , who in fact prove the exact bound $N\geq \binom{n}{\lfloor n/2\rfloor}$.
+- In [Er73] and [ErGr80] the generalisation where $A\subseteq (0,N]$ is a set of real numbers such that the subset sums all differ by at least $1$ is proposed, with the same conjectured bound. (The second proof of [DFX21] applies also to this generalisation.) This generalisation seems to have first appeared in [Gr71] .
+
+Public-site preview:
+- OPEN
+- This is open, and cannot be resolved with a finite computation.
+- - $500
+- If $A\subseteq \{1,\ldots,N\}$ with $\lvert A\rvert=n$ is such that the subset sums $\sum_{a\in S}a$ are distinct for all $S\subseteq A$ then\[N \gg 2^{n}.\]
+- #1 : [Er56] [Er57] [Er59] [Er61] [Er65b] [Er69] [Er70b] [Er70c] [Er73] [BeEr74,p.619] [ErSp74] [Er75b] [Er77c] [ErGr80,p.59] [Er81] [Er82e] [Er85c] [Er90] [Er91] [Er92b] [Er95,p.165] [Er97c] [Er98] [Va99,1.20]
+- number theory |

package/problems/1/problem.yaml

@@ -0,0 +1,44 @@
+problem_id: "1"
+display_name: "Erdos Problem #1"
+title: "Erdos Problem #1"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/1
+  external_id: "1"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "1"
+provenance:
+  seeded_at: 2026-03-26T04:41:24.787Z
+  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: $500
+related_problems: []
+family_tags:
+  - number-theory
+  - additive-combinatorics
+harness:
+  depth: dossier
+statement:
+  short: If $A\subseteq \{1,\ldots,N\}$ with $\lvert A\rvert=n$ is such that the
+    subset sums $\sum_{a\in S}a$ are distinct for all $S\subseteq A$ then\[N \gg
+    2^{n}.\]
+  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/19/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 19`
+- `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 19 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/19/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 19
+- 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/19/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #19 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/19/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/19/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/19>
+- 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/19/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 19: `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/19/STATEMENT.md

@@ -0,0 +1,15 @@
+# Problem 19 Statement
+
+Source: <https://www.erdosproblems.com/19>
+
+Normalized focus:
+- Curated dossier seeded from upstream public metadata for Erdos Problem #19.
+- Seeded with preview lines from the public site snapshot
+
+Public-site preview:
+- Erdős Problem #19
+- -
+- -
+- Forum
+- Inbox
+- Favourites

package/problems/19/problem.yaml

@@ -0,0 +1,42 @@
+problem_id: "19"
+display_name: "Erdos Problem #19"
+title: "Erdos Problem #19"
+source:
+  site: erdosproblems.com
+  url: https://www.erdosproblems.com/19
+  external_id: "19"
+upstream:
+  repo: https://github.com/teorth/erdosproblems
+  data_file: data/problems.yaml
+  number: "19"
+provenance:
+  seeded_at: 2026-03-26T04:41:26.625Z
+  seeded_from:
+    kind: pull_bundle
+    upstream_record_included: true
+    site_snapshot_included: true
+status:
+  site_status: decidable
+  site_badge: DECIDABLE
+  repo_status: cataloged
+  upstream_status: decidable
+  upstream_last_update: 2025-08-31
+cluster: graph-theory
+prize:
+  display: $500
+related_problems: []
+family_tags:
+  - graph-theory
+  - chromatic-number
+harness:
+  depth: dossier
+statement:
+  short: "Curated dossier seeded from upstream public metadata for Erdos Problem #19."
+  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/2/AGENT_START.md

@@ -0,0 +1,17 @@
+# Agent Start
+
+Fast start:
+- `erdos problem show 2`
+- `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 2 against its pull bundle, references, and upstream provenance before widening claims.

package/problems/2/CHECKPOINT_NOTES.md

@@ -0,0 +1,11 @@
+# Checkpoint Notes
+
+- Problem: 2
+- 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/2/EVIDENCE.md

@@ -0,0 +1,10 @@
+# Evidence
+
+- This dossier was seeded for Erdos Problem #2 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/2/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/2/REFERENCES.md

@@ -0,0 +1,9 @@
+# References
+
+- Public problem page:
+  - <https://www.erdosproblems.com/2>
+- 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/2/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 2: `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/2/STATEMENT.md

@@ -0,0 +1,15 @@
+# Problem 2 Statement
+
+Source: <https://www.erdosproblems.com/2>
+
+Normalized focus:
+- Curated dossier seeded from upstream public metadata for Erdos Problem #2.
+- Seeded with preview lines from the public site snapshot
+
+Public-site preview:
+- Erdős Problem #2
+- -
+- -
+- Forum
+- Inbox
+- Favourites