Evidence Record: Templar Protocol (NEAR), Immunefi #77956
Supporting evidence for Templar Protocol (NEAR): Keyless Withdrawal-Queue Freeze & Unpaid $100K Ruling. This page preserves the full deployment census, keyless-market accounting, disclosure chronology, and verification commands without interrupting the technical narrative.
Timeline dates are shown as recorded in the source screenshots and exports; clock times are omitted there. On-chain block heights remain the authoritative record.
1. Deployment record: 22 keyed markets
The census below was taken at block 218,233,933 on October 2, 2026, before the October 7–8 redeployments. Values are the displayed, converted amounts from the investigation record. The aggregate reported in that source is $15,402,321.89 (approximately $15.40M). Individual rounded rows do not add to that aggregate exactly.
Mainnet metadata labels the deployed bytecode v1.6.0, code hash 7WUFT7JR2mbKhQkPNedLL3ZZ28arqbwAFFubaMQpRpPW, NEP-330 revision 07e20fc7. That revision was not found as a public commit. The public bounds patch is PR #659, commit dbfcd42b. This evidence does not establish that the deployed WASM is byte-identical to that PR.
| # | Market | Supply at Oct 2 | Deployed (UTC) | Links |
|---|---|---|---|---|
| 1 | iethwbtc-ixlmusdc.v1.tmplr.near | $5,167,284.41 | 2026-10-08 14:42 | Contract | Deploy Tx |
| 2 | ibtc-ixlmusdc.v1.tmplr.near | $3,408,420.24 | 2026-10-08 14:39 | Contract | Deploy Tx |
| 3 | ixlmdejaaa-ixlmusdc-1.v1.tmplr.near | $3,239,143.57 | 2026-10-08 14:12 | Contract | Deploy Tx |
| 4 | ixlm-ixlmusdc-1.v1.tmplr.near | $1,441,662.99 | 2026-10-08 14:37 | Contract | Deploy Tx |
| 5 | ibtc-iethusdc-1.v1.tmplr.near | $1,042,748.46 | 2026-10-08 14:39 | Contract | Deploy Tx |
| 6 | ixlmdejaaa-ixlmusdc-2.v1.tmplr.near | $400,000.90 | 2026-10-08 14:13 | Contract | Deploy Tx |
| 7 | ixlmsolvbtc-ixlmusdc.v1.tmplr.near | $100,005.10 | 2026-10-08 14:30 | Contract | Deploy Tx |
| 8 | ixrp-ixlmusdc.v1.tmplr.near | $100,003.35 | 2026-10-08 14:32 | Contract | Deploy Tx |
| 9 | izec-ixlmusdc.v1.tmplr.near | $100,000.14 | 2026-10-08 14:33 | Contract | Deploy Tx |
| 10 | ixlmcetes-ixlmusdc.v1.tmplr.near | $100,000.11 | 2026-10-08 14:14 | Contract | Deploy Tx |
| 11 | ixlmustry-ixlmusdc.v1.tmplr.near | $100,000.05 | 2026-10-08 14:31 | Contract | Deploy Tx |
| 12 | iada-ixlmusdc.v1.tmplr.near | $100,000.04 | 2026-10-08 14:37 | Contract | Deploy Tx |
| 13 | idoge-ixlmusdc.v1.tmplr.near | $100,000.04 | 2026-10-08 14:40 | Contract | Deploy Tx |
| 14 | iethhemibtc-iethusdc.v1.tmplr.near | $1,031.60 | 2026-10-08 14:41 | Contract | Deploy Tx |
| 15 | iethfxrp-ixlmusdc.v1.tmplr.near | $1,001.20 | 2026-10-08 14:40 | Contract | Deploy Tx |
| 16 | iltc-ixlmusdc.v1.tmplr.near | $1,001.00 | 2026-10-08 14:42 | Contract | Deploy Tx |
| 17 | ixlmdejtrsy-ixlmusdc.v1.tmplr.near | $5.82 | 2026-10-08 14:11 | Contract | Deploy Tx |
| 18 | ixlmdejaaa-ixlmusdc.v1.tmplr.near | $5.26 | 2026-10-08 14:10 | Contract | Deploy Tx |
| 19 | linear-usdt.v1.tmplr.near | 3.50 USDT | 2026-10-07 22:24 | Contract | Deploy Tx |
| 20 | stnear-usdt.v1.tmplr.near | 3.00 USDT | 2026-10-08 14:12 | Contract | Deploy Tx |
| 21 | ixlm-ixlmpyusd.v1.tmplr.near | 2.16 PYUSD | 2026-10-08 14:43 | Contract | Deploy Tx |
| 22 | ixlmdejtrsy-ixlmusdc-1.v1.tmplr.near | $0.00 | 2026-10-08 14:29 | Contract | Deploy Tx |
The two keyless accounts discussed in the main article were not part of these 22 deployments.
2. Keyless market: July migration and residual positions
On January 19, 2026, block 181,696,954, ixlm-ixlmusdc.v1.tmplr.near deleted ed25519:5gEdy3qyZBZveGfcxcUTGCoVzzKnZXaQASWjmyP44giZ in 5PPCjn583spZKptMEo1gYN4zuhCRhnWKwsTKxag3FQFd. Immunefi counted 114 days between that deletion and the May 13 filing.
The successor ixlm-ixlmusdc-1.v1.tmplr.near was deployed on July 14, 2026 by v1.tmplr.near in F4Eh6Bo6A1ufY6DLGXdHgYmJBXvHdrd5JdKc1mTGjBRU. It retained a full-access key and changed the withdrawal minimum from 0.04 to 1.00 USDC.
Both withdrawals were signed by the implicit account 80635cbbea3bc03a112c72a30ad94059213a253db65f549c8c406e707d730054:
| When (UTC) | Action | USDC | Block / transaction |
|---|---|---|---|
| Jul 16, 20:15 | Withdraw from keyless market | 1,084,536.1225777 | 207,255,140 / Tx |
| Jul 16, 20:27 | Supply to successor | 1,084,536.12 | 207,255,899 / Tx |
| Jul 27, 17:00 | Withdraw from keyless market | 357,091.0044225 | 208,788,406 / Tx |
| Jul 27, 17:04 | Supply to successor | 357,091.00 | 208,788,791 / Tx |
The exact withdrawals total 1,441,627.1270002 USDC. The displayed successor deposits total 1,441,627.12 USDC, approximately 96 percent of the submission-time supply used by mediation.
Funding and residual positions
Three records establish an initialization and gas-funding path:
- Block 171,233,785,
3D76DGqzaGhLvV6fdhqQ6FyoZHTwK1Q6ue65tkt7ortN:templarllc.neartransferred 2.0 NEAR tosimrat.near. - Block 172,489,440,
AFyWbs2ysyQJTSfwNEYXfDaQLRhm3rMSrLYKgLk8nYtv:team.herewallet.nearinitialized80635cb…. - Block 172,489,521,
EiYoieskajddQPRMU68eHp2yFffG8ZaHamHopdaUUg1Z:simrat.neartransferred 0.8 NEAR to80635cb….
80635cb… then supplied 1,500,000 USDC in EgDkqFBYyUzX9VxGB6kyJD2fWzv9aK1fZufx5X145p6q. These records establish a funding path, not beneficial ownership, present common key control, or who authorized the July transactions.
At block 219,400,038, and again around blocks 223,518,994–223,519,017, the keyless market had 59,137.9738191 USDC active supply. At the later snapshot, the borrowed amount was 18,090.1807530 USDC, the borrow-side available metric was 40,456.4133279 USDC, and the queue length was 0. Listed supply positions included:
80635cb…: 58,372.8818437 USDC active, plus 853.6169686 USDC accrued yield.simrat.near: 755.60 USDC active.- The other 13 listed accounts: approximately 2.3 USDC active, plus small incoming amounts.
At the July 27 snapshot, 54,985.25 USDC, approximately 93 percent of the pool, was borrowed. Repayments and one liquidation returned 37,079.75 USDC between July 30 and August 20, reducing active borrows to about 18,090.18 USDC and increasing liquid assets. Neither the historical repayment figure nor the borrow-side available metric guarantees a particular future withdrawal.
3. Disclosure timeline
Dates below are shown without clock times to avoid conflating dashboard display time with UTC on-chain time. The underlying screenshots retain their original timestamps.
- May 13, 2026. #77956 filed and escalated with Immunefi. The initial filing focused on v1.3.0 and the keyed
ibtc-iethusdc-1.v1.tmplr.nearmarket. - May 16, 2026. An update titled “Extended Impact Scope: $4.75M at Risk Across 18 Markets” expanded the analysis to v1.2.0, v1.2.1, and v1.3.0 and included the keyless XLM/USDC market at approximately $1.50M.
- May 25, 2026. Mediation requested.
- June 8, 2026. Templar’s public migration map listed
ixlm-ixlmusdcas TBD. - June 24, 2026. On the ticket, severity was moved from Critical to Medium on a griefing theory.
- June 29, 2026. Templar wrote that it was evaluating whether a fix was necessary. The keyless XLM/USDC and isolated-ZEC markets had zero access keys.
- July 2, 2026. Immunefi paused Templar’s program for unresponsiveness.
- July 8–9, 2026. Direct warning to RoyalF00l about the keyless market. The July 9 reply is quoted in the main article.
- July 14, 16, and 27, 2026. Keyed successor deployed; approximately $1.44M USDC moved. The July 15 migration map named the successor.
- August 7, 2026. Immunefi removed Templar’s program for repeated SLA failures.
- September 13, 2026. An Immunefi staff DM initially said publication was possible because the project was not responding.
- September 15, 2026. Commit
6e561837changedSECURITY.md: “Templar’s former Immunefi bug bounty program has ended; please do not submit reports there.” - September 22, 2026. I sent Templar a formal public-disclosure warning.
- September 26, 2026. RoyalF00l wrote that Certora had verified the findings and fix and requested a Medium-tier invoice. I rejected Medium and restated the keyless market’s permanence.
- September 28, 2026. RoyalF00l accepted that the finding was valid but argued Medium/Griefing, the 1,000 TGas limit, and L-05.
- September 29–30, 2026. Immunefi requested a full-gas local-fork test. The September 30 run hit the 4 MB per-receipt recorded-storage-proof limit at 734.8 TGas.
- October 1, 2026. Mediation corrected the publication record: confidentiality had not been released.
- October 2, 2026. Templar confirmed the mechanism, according to the final determination.
- October 7–8, 2026. Twenty-two keyed markets received bytecode labeled v1.6.0.
- October 8, 2026. RoyalF00l argued the May proof used a superseded 300 TGas limit, cited L-05, and offered $2,500 as a goodwill payment.
- October 9, 2026. I declined the $2,500 and cited
a96e44d,ac072f1f, and08d32ef. - October 22, 2026. Final determination: Critical, $100,000, with Griefing, L-05, and Sybil rejected. Category 2 publication rights applied.
- October 26, 2026. I sent the determination and repeated the safety note. No reply is in that thread.
- November 2, 2026. I offered a $60,000 settlement, open until November 8. The email referred to the keyless market but did not repeat the residual-supplier withdrawal warning.
- November 11, 2026 (UTC) / November 11–12 (local PDF print).
@RoyalF00loffered $3,000 (15:44 UTC) citingL-05and then (20:17 UTC) conditioned any payment on permanent non-publication of this write-up.
Immunefi’s July 2 pause and August 7 removal notices, followed by Templar’s September 15 change to SECURITY.md. Discord crops exclude unrelated reports and participants.
4. Independent verification
These commands inspect public state or local Git history. Output changes with chain state; use block-specific queries when comparing a historical snapshot.
Check both keyless accounts:
1
2
near account list-keys ixlm-ixlmusdc.v1.tmplr.near network-config mainnet now
near account list-keys izec-isolusdc.v1.tmplr.near network-config mainnet now
Inspect supply, borrowing, and the borrow-side available metric:
1
2
3
4
5
6
7
near contract call-function as-read-only \
ixlm-ixlmusdc.v1.tmplr.near get_borrow_asset_metrics \
json-args '{}' network-config mainnet now
near contract call-function as-read-only \
izec-isolusdc.v1.tmplr.near get_borrow_asset_metrics \
json-args '{}' network-config mainnet now
Check the withdrawal queues:
1
2
3
4
5
6
7
near contract call-function as-read-only \
ixlm-ixlmusdc.v1.tmplr.near get_supply_withdrawal_queue_status \
json-args '{}' network-config mainnet now
near contract call-function as-read-only \
izec-isolusdc.v1.tmplr.near get_supply_withdrawal_queue_status \
json-args '{}' network-config mainnet now
Check the deployed market metadata:
1
2
3
near contract call-function as-read-only \
ixlm-ixlmusdc.v1.tmplr.near contract_source_metadata \
json-args '{}' network-config mainnet now
For the Git history, work inside a clone of Templar-Protocol/contracts. The public bounds patch was squash-merged, so fetch the PR head before querying dbfcd42b:
1
2
3
4
5
6
git fetch origin pull/659/head
git show -s --format='%H | author %aI | commit %cI | %s' \
a96e44dc9ed34f3251975a06a709ea94bd944922 \
ac072f1fa475ef644f99e4ccffdbfded6d2b87c4 \
08d32ef177fee011badb72766be79ddb085dd287 \
dbfcd42b4ee5565bffc05e652c07af463a2bdd19
5. Primary sources
- Immunefi mediation determination, October 22, 2026, SHA-256
78e3e280bfa778f00233e6d27e68cc6415ec9e6a54db8f6c84bd491b70ad63fa. - Certora, Security Assessment and Formal Verification Report, Templar Protocol, September 2025. Scope on page 4; L-05 on page 86.
- Vulnerable revision
e1f5b676; keyless source revision877cdf79; public bounds patch PR #659. - NEP-509; nearcore 2.11.0 witness-size parameters and Protocol 83 gas parameters.