Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add synaptiai/synapti-marketplace --skill finding-reconciliationgit clone --depth 1 https://github.com/synaptiai/synapti-marketplaceWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/synaptiai/synapti-marketplace/finding-reconciliation)<a href="https://agentmods.dev/skills/synaptiai/synapti-marketplace/finding-reconciliation"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/finding-reconciliation/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/synaptiai/synapti-marketplace/finding-reconciliation"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/finding-reconciliation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00150 | $0.01979 |
| Opus 5 | $0.00075 | $0.00989 |
| Sonnet 5 | $0.00030 | $0.00396 |
| Haiku 4.5 | $0.00015 | $0.00198 |
Grade A, and why
finding-reconciliation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Finding Reconciliation
Phase 9. Turns three independent findings tables into one adjudicated ledger, then drives repair and re-audit.
Iron Law
PUBLISH THE FINDINGS TABLE BEFORE ANY REPAIR, AND RE-AUDIT IN FRESH CONTEXTS AFTER. A finding seen by one pass is still a finding.
A package that quietly fixes what it found and reports only the clean end state has destroyed its own audit trail. The pre-repair table is the evidence that verification happened.
Never loaded by a verifier
This skill must not appear in any pass agent's skills: list. If a verifier could load the merge logic, it would reason about corroboration and drift toward the consensus it expects the other passes to reach — which is the correlated-error failure the whole three-pass architecture exists to prevent.
tests/agent-independence.test.sh enforces this mechanically. It is a structural guarantee, not a convention.
Order of operations
- Collect raw. Read each pass's table from
.dossier/runs/<id>/pass-{A,B,C}.mdverbatim. Do not edit, do not summarize, do not reorder. - Normalize to the canonical field set.
- Deduplicate by
(location_normalized, claim_hash). Keep the highest severity. Merge the problem descriptions when they add different context — do not discard the second one's reasoning. - Stamp corroboration as
N/3. - Detect disagreement and promote it.
- Publish the pre-repair table. This is a commit point, not a draft.
- Split into agent-repairable and
needs-owner. - Apply the repairable ones, Critical and High first.
- Re-run every check a correction touched.
- Re-audit in fresh contexts when a correction materially changed architecture, product behaviour, data handling, security, reliability, or a public claim.
Corroboration does not rank
3/3 means three lenses independently saw it. 1/3 means one lens looked in the right place. Neither is a confidence score.
A 1/3 finding is never downgraded for being lonely. The passes have different lenses by design — pass C is the only one simulating a new engineer's first contribution, so a broken onboarding command should be 1/3, and it is no less real for that. Downgrading by vote count converts three specialized reviewers into a majority-vote committee and throws away the specialization.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 126 lines · 150 tokens per session scan A 12d643def162
finding-reconciliation is a skill published in the GitHub repository synaptiai/synapti-marketplace (6 stars, last pushed yesterday), licensed Apache-2.0. It adds 150 tokens to every session and 1,979 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
step4-archive-spec
Archive a completed spec — moves it to specs-archived and blocks agent access.
plan-roadmap
Create or update a prioritized product roadmap with phased milestones.
harvest-findings
Composable operation that harvests inline manual-test findings from an epic's test plan, captures them to the backlog, and turns in-scope ones into a remediation story. Called at the manual-test-pass boundary by /epic-flywheel and /retrospective. Directly invocable as /harvest-findings {N}.
debug
Run /debug to find and fix a bug's root cause: a test failing for an unclear reason, /check verify finding a failure, or behavior being wrong. Runs a reproduce, localize, hypothesize, test, fix, verify loop, makes the minimal fix, and hands a regression test to /test. No features, no extra refactors.
check
Confirm a change before merge. /check verify drives the real app to prove behavior against the spec (every acceptance criterion met, every surface built). /check review runs a senior code review on a fresh model, one that did not write the code. Verify after /develop, review before a PR. Writes to docs/reviews/, never…
pipeline-efficiency-benchmark
A repeatable benchmark for comparing different versions of a development pipeline. It runs the same public tasks in isolated workspaces and records scores, tokens, time, tool calls, and cost.