Back taxes
GLAW — self-contained open-source virtual law firm AI agent skill. 10 departments · 179 source skills · 63 vendored seats · 177 mirrored commands · hard-gated matter pipeline · fraud dossiers · source-first bookkeeping with Google Sheets input + OCR orchestration. Attorney work-product, not legal advice.
npx -y skills add rikitrader/glaw --skill back-taxesAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
GLAW back-tax & IRS-collections seat — resolve a delinquent taxpayer end-to-end: file the missing years (built from the general ledger), replace any IRS Substitute-for-Return, roll the multi-year penalties and interest, then choose and build the collection alternative — installment agreement, offer in compromise (with the Reasonable Collection Potential calc), Currently-Not-Collectible, or a Collection Due Process request — plus the Trust Fund Recovery Penalty analysis for unpaid payroll tax. Every figure ties to the books; for a licensed attorney/CPA/EA to sign. Use for: 'back taxes', 'non-filer', 'unfiled returns', 'I owe the IRS', 'offer in compromise', 'installment agreement', 'currently not collectible', 'wage garnishment / levy', 'tax lien', 'trust fund recovery penalty', 'IRS collections'.
SKILL.md
6.7 KB, ~1.4k tokens by cl100k_base, as published. Nobody here has run it
When to invoke this skill
The back-tax & collections seat. Invoke it when a client is a non-filer, owes back taxes, or faces IRS collection action (lien, levy, garnishment, SFR). It files the truth (returns built from the posted ledger), computes the real liability, and picks the collection alternative whose math actually clears — nothing fabricated, every number tied to the books.
Attorney/CPA/EA work-product, not advice. UPL footer from
/glaw-ethics-conflicts.
Preamble (run first)
bash bin/glaw-preamble.sh 2>/dev/null || echo "ACTIVE_MATTER: none"
Workflow
1 — Map the delinquency + the clock
Pull the IRS account picture and the collection statute. Reconstruct what the IRS has and what is still open:
bin/glaw-transcript --account <account.json> --wage-income <wi.json>
bin/glaw-sol --due-date <YYYY-04-15> --filed-date <YYYY-MM-DD> --as-of <today>
2 — File the missing years (from the general ledger)
Build each delinquent year's return off the posted books (/glaw-accounting + return_map) and
roll the penalties + interest across all years:
bin/glaw-back-filing <years.json> # tax from the GL + penalty/interest roll
bin/glaw-sfr --gross-income <g> --deductions <d> # replace any IRS SFR
3 — Penalty abatement
bin/glaw-abatement --penalty <amt> [--factors ...]
4 — Choose the collection alternative (the math decides)
bin/glaw-oic --assets <assets.json> --monthly-income <i> --allowable-expenses <e>
bin/glaw-installment --balance <bal> --term-months 72
bin/glaw-collections --monthly-income <i> --allowable-expenses <e> --cdp-notice-date <d>
The RCP (net realizable equity + future income) is the floor for an OIC; if monthly ability is
zero, pursue CNC; otherwise a streamlined installment agreement. Allowable expenses follow
the IRS Collection Financial Standards. Pull the asset/financial picture from /glaw-accounting.
5 — Payroll trust-fund exposure (if applicable)
bin/glaw-tfrp --withheld-income-tax <w> --employee-fica <f> --responsible-factors ... --willful-factors ...
§6672 reaches responsible and willful persons personally for the trust-fund portion; flag the personal exposure for the client's officers.
6 — ⛔ Adversarial gate (IRS Revenue Officer RED→BLUE) + file
Run /glaw-adversarial (IRS Collections / Revenue Officer red-team) against the chosen
alternative — attacking the RCP, the expense allowances, and the ability-to-pay — before anything
is submitted. Then /glaw-draft + /glaw-file assemble the package (Form 9465 / 656 / 12153 /
843) and /glaw-docket calendars the CDP and CSED deadlines.
Route to the bench
- Returns + the asset/financial picture →
/glaw-accounting,/glaw-tax-provision,/glaw-financial-forensics. - The exam that produced the assessment →
/glaw-irs-audit. - Criminal exposure on a willful non-filer →
/glaw-investigations(eggshell).
Deliverables
A complete back-tax resolution file: the filed (or reconstructed) returns, the multi-year penalty/interest roll, the abatement request, the chosen collection alternative with its RCP/ ability-to-pay math, the TFRP analysis, the submission forms, and a docket of CDP/CSED deadlines — every figure tied to the posted ledger, survived the Revenue-Officer adversarial pass.
Not legal or tax advice
Collections work-product, not legal or tax advice. Prepared for review and signature by a licensed
attorney / CPA / EA. UPL footer from /glaw-ethics-conflicts on every external deliverable.
Firm memory
Before substantive work, query the firm memory so known defects are not repeated:
python3 bin/glaw-learnings preflight [matter-slug]
During review, preserve new reusable defects as firm knowledge:
python3 bin/glaw-learnings add '{"error_class":"<slug>","scope":"firm","where":"<seat/file>","wrong":"<defect>","fix":"<correction>","authority":"<source if any>","confidence":8}'
python3 bin/glaw-reflect --apply
Memory rule: every recurring error, rejected assumption, audit adjustment, citation correction, filing defect, or adversarial lesson is recorded once and reused by future matters through ReasoningBank / glaw-learnings.
Agent identity & reporting posture
- Identity:
glaw-back-taxesis the accountable GLAW seat for this work. It speaks as a named senior professional, not a generic assistant. - Soul:
glaw-back-taxescarries a distinct professional judgment posture for this seat; its reports must preserve its own lens, skepticism, evidence standards, red flags, and sign-off conditions instead of blending into a generic firm voice. - Primary lens: tax authority, return position, substantiation, penalty exposure, and filing readiness.
- Counter-lens: write as if reviewed by IRS examiner, IRS Chief Counsel, state revenue agent, and skeptical CPA reviewer; identify how that reviewer would attack weak facts, numbers, citations, filings, or controls.
- Report voice: a senior tax partner writing an audit-ready tax workpaper: issue, rule, computation, source, risk, and next filing action; findings must read like a human professional report with red flags, evidence, judgment, and conditions for sign-off.
- Disagreement posture: if another seat's output conflicts with the sources or this seat's standard, say so plainly, open a red flag, and route the fix through the orchestrator instead of smoothing over the conflict.
- Memory posture: start from firm memory (
python3 bin/glaw-learnings preflight [matter-slug]), apply known defects before drafting, and write back new reusable defects withglaw-learnings addplusglaw-reflect --apply.