Context
Following an extensive multi-agent research and gap-analysis pass (2026-07-08), cross-checked
directly against the live issue tracker, not just code. This is deliberately the complement
of the existing, already ~80-85% complete roadmap (#2589 master tracker, Wave 1-4, #2111,
#2112, #2552, #2565, #3357-#3365 and others) — every epic below was checked against the
tracker before filing to avoid duplicates.
Key correction from earlier assumptions this session: current per-neuron metagraph state is
already captured and served (.github/workflows/refresh-metagraph.yml -> scripts/fetch-metagraph-native.py
-> D1 neurons -> 25+ live routes). That is NOT part of this tree. The one large, confirmed,
unfiled capture gap is subnet hyperparameters (see #EPIC1) — everything else here is either
a derived view on our already-verified-accurate log layer (blocks/extrinsics/events), or a
narrower enrichment of an existing route.
See docs/block-explorer-data-model.md for the full technical reference this tree is built on
(extrinsics vs events vs state, the archive RPC surface, the competitor feature benchmark).
Scope fence — respected, not decided here
#2589 explicitly excludes trader/wallet/tax features (portfolio/PnL, OHLC candlesticks, wallet
actions). Two items from the competitor benchmark brush against that fence and are flagged,
not epic'd — see the dedicated decision issue linked below. Do not build them without an
explicit maintainer call.
Deliverables
9 epics, each linked below as a sub-issue with its own contributor-sized sub-tasks (context,
deliverable, and which existing pattern to follow, cited precisely). Most sub-tasks are
startable immediately against current chain state; a few are explicitly blocked on the archive
node reaching tip (~2026-07-20) and are labeled as such in their own issue body.
Expected outcome
Closes the confirmed gap between metagraphed's current data surface and full feature parity
with taostats.io/taomarketcap.com (within the developer-explorer scope fence), leaving a clean,
independently-pickable backlog ready to relabel as gittensor:feature contributor bounties.
maintainer-only — planning/tracking issue, not itself a build task.
Context
Following an extensive multi-agent research and gap-analysis pass (2026-07-08), cross-checked
directly against the live issue tracker, not just code. This is deliberately the complement
of the existing, already ~80-85% complete roadmap (#2589 master tracker, Wave 1-4, #2111,
#2112, #2552, #2565, #3357-#3365 and others) — every epic below was checked against the
tracker before filing to avoid duplicates.
Key correction from earlier assumptions this session: current per-neuron metagraph state is
already captured and served (
.github/workflows/refresh-metagraph.yml->scripts/fetch-metagraph-native.py-> D1
neurons-> 25+ live routes). That is NOT part of this tree. The one large, confirmed,unfiled capture gap is subnet hyperparameters (see #EPIC1) — everything else here is either
a derived view on our already-verified-accurate log layer (blocks/extrinsics/events), or a
narrower enrichment of an existing route.
See
docs/block-explorer-data-model.mdfor the full technical reference this tree is built on(extrinsics vs events vs state, the archive RPC surface, the competitor feature benchmark).
Scope fence — respected, not decided here
#2589 explicitly excludes trader/wallet/tax features (portfolio/PnL, OHLC candlesticks, wallet
actions). Two items from the competitor benchmark brush against that fence and are flagged,
not epic'd — see the dedicated decision issue linked below. Do not build them without an
explicit maintainer call.
Deliverables
9 epics, each linked below as a sub-issue with its own contributor-sized sub-tasks (context,
deliverable, and which existing pattern to follow, cited precisely). Most sub-tasks are
startable immediately against current chain state; a few are explicitly blocked on the archive
node reaching tip (~2026-07-20) and are labeled as such in their own issue body.
Expected outcome
Closes the confirmed gap between metagraphed's current data surface and full feature parity
with taostats.io/taomarketcap.com (within the developer-explorer scope fence), leaving a clean,
independently-pickable backlog ready to relabel as
gittensor:featurecontributor bounties.maintainer-only — planning/tracking issue, not itself a build task.