Current RNA editing techniques are predominantly limited to single-base edits. Here, we introduce selective cleavages and intramolecular stitches of RNA (SCISSOR) for selective cleavage and intramolecular stitches of RNA. Building on the principle that type III CRISPR complex determines target cleavage positions based on gRNA length in 6-nt increments, we hypothesized that engineering gRNAs with bulge loops could circumvent this rule, allowing for flexible RNA excision. Through systematic evaluation of gRNAs with various bulge loops, we established the rules for precise non-6-nt target cleavage and repair. We observed that the complex tolerates 1- or 2-nt bulge loops and accommodates large bulge loops ranging from 6 to 24 nt. Consequently, SCISSOR could accomplish nearly any length of short fragment excision. With its capability to modify open reading frames, we demonstrate the potential of SCISSOR in repairing frameshift mutations and introducing frameshifts to create immunogenic poly-epitopes in human cells. SCISSOR holds promise in RNA therapy and biomedical research.